docs: correct scene-boundary detector F1 numbers to measured values

The boundary-detection paragraph understated the detector. Replace the stale
"~34% F1 vs ~27%" (a pre-C++-retrain figure with no backing artifact) with the
measured numbers at the shipped ±20s tolerance:
  - leave-one-out macro boundary F1 = 44.1% (honest generalisation)
  - grayscale baseline               = 29.8%
  - train-all (shipped model)        = 72.9% (per-film 51-86%)
computed from experiments/results/scene_boundary/xgb_report.json and per-film
leave-one-out runs of train_xgb_cpp.py. Also correct the false claim that the
low-contrast grades "cannot generalise held out" — Scarface held out scores 32%,
Café Society 51%, both above grayscale (0% and 31%).

Split the evolution figure into two panels so the strict-±2s feature-development
curve is no longer mistaken for the shipped result: left = feature progress at
±2s, right = shipped detector at the ±20s tolerance the pipeline uses.

.gitignore

@@ -1,5 +1,6 @@
 # Build
 build/
+build-*/
 cmake-build-*/
 CMakeCache.txt
 CMakeFiles/
@@ -117,3 +118,6 @@ venv/
 *.swo
 .DS_Store
 Thumbs.db
+.venv-rocm/
+!models/scene_boundary_xgb.json
+experiments/dump_review/

CMakeLists.txt

@@ -280,6 +280,24 @@ FetchContent_Declare(
 )
 FetchContent_MakeAvailable(nanobind)
 
+# XGBoost (learned scene-boundary detector for flood-fill presence). Fetched and
+# built from source so we get both the C API header and a matching libxgboost,
+# reproducibly — the pip wheel ships the .so but no header. Heavy first build, so
+# it is opt-in; the scene-boundary node is compiled only when SAE_SCENE_XGB is on.
+option(SAE_SCENE_XGB "Build the XGBoost scene-boundary detector node" ON)
+if(SAE_SCENE_XGB)
+    set(BUILD_STATIC_LIB ON  CACHE BOOL "" FORCE)   # link xgboost statically
+    set(USE_OPENMP       ON  CACHE BOOL "" FORCE)
+    FetchContent_Declare(
+        xgboost
+        GIT_REPOSITORY https://github.com/dmlc/xgboost.git
+        GIT_TAG        v2.1.1
+        GIT_SHALLOW    TRUE
+        GIT_SUBMODULES_RECURSE TRUE
+    )
+    FetchContent_MakeAvailable(xgboost)
+endif()
+
 # ── Model paths ───────────────────────────────────────────────────────────────
 set(SAE_MODELS_DIR "${CMAKE_SOURCE_DIR}/models"
     CACHE PATH "Directory containing ONNX model files")
@@ -326,32 +344,17 @@ target_link_libraries(sae_embed PRIVATE sae_gallery)
 # network (KPN_BUILD_PYTHON is enabled per-TU inside the .cpp). Powers the
 # threshold-sweep optimizer in scripts/optimizer/.
 #
-# OFF by default, and this is a statement of fact rather than a preference: the
-# module HAS NOT COMPILED since the AR-007/AR-008 tracker redesign. FaceTrackerFunc
-# now requires a TrackRegistry and a calibration at construction, and the binding
-# still builds it from a Config alone. The .so in a stale build/ directory
-# predates that change.
-#
-# Fixing it is VR-011's job, not a patch: the tracker needs the calibration, the
-# calibration comes from the matcher, and the matcher is added to the network
-# afterwards -- so the seam has to be restructured, exactly as main.cpp already
-# is (matcher first, then registry, then tracker). Presence claims do not cross
-# the seam at all today, which is the other half of the same rewrite.
-#
-# Recorded as a switch rather than left as a build error so that `cmake --build`
-# succeeds and the breakage is attributed instead of rediscovered. Turning it on
-# reproduces the failure immediately, which is the point.
-#
 # TRACES: VR-011 | PR-002
-option(SAE_BUILD_KPN_BINDINGS
+# ON again. It was OFF for one commit because it had not compiled since the
-       "Build the sae_kpn Python module (BROKEN pending VR-011)" OFF)
+# AR-007/AR-008 tracker redesign -- the binding built FaceTrackerFunc from a
+# Config alone, and the tracker had required a registry and a calibration since.
+# VR-011 replaced the three per-node factories with one `add_pipeline` that
+# builds the chain in main.cpp's order, which is the only order that satisfies
+# those dependencies, so the failure mode cannot recur from Python.
+option(SAE_BUILD_KPN_BINDINGS "Build the sae_kpn Python module" ON)
 if(SAE_BUILD_KPN_BINDINGS)
     nanobind_add_module(sae_kpn src/kpn_bindings.cpp)
     target_link_libraries(sae_kpn PRIVATE sae_gallery)
-else()
-    message(STATUS
-        "sae_kpn: SKIPPED (SAE_BUILD_KPN_BINDINGS=OFF). The Python replay "
-        "bindings do not compile against the post-AR-012 tracker; see VR-011.")
 endif()
 
 # ── sae_audio — Python module: the v1 audio signature (IR-004) ────────────────
@@ -367,16 +370,43 @@ target_link_libraries(sae_audio PRIVATE ffmpeg_libs)
 # HDF5 already found above (before sae_gallery); vars HDF5_CXX_LIBRARIES / _INCLUDE_DIRS
 # are reused by scene_analyze / dump_embeddings below.
 
+# The learned scene-boundary detector is compiled into the sink (result_sink →
+# xgb_scene_boundary + audio_logpsd) when SAE_SCENE_XGB is on, so the analysis
+# binaries need xgboost + FFTW + ffmpeg and the define. Found once here.
+if(SAE_SCENE_XGB)
+    find_library(FFTW3_LIB fftw3 REQUIRED)
+    set(SAE_SCENE_LIBS xgboost ${FFTW3_LIB} ffmpeg_libs)
+    set(SAE_SCENE_DEFS SAE_SCENE_XGB)
+else()
+    set(SAE_SCENE_LIBS "")
+    set(SAE_SCENE_DEFS "")
+endif()
+
 # ── analyze — main analysis binary ───────────────────────────────────────────
 add_executable(scene_analyze src/main.cpp)
-target_link_libraries(scene_analyze PRIVATE sae_gallery ${HDF5_CXX_LIBRARIES})
+target_link_libraries(scene_analyze PRIVATE sae_gallery ${HDF5_CXX_LIBRARIES} ${SAE_SCENE_LIBS})
 target_include_directories(scene_analyze PRIVATE ${HDF5_INCLUDE_DIRS})
+target_compile_definitions(scene_analyze PRIVATE ${SAE_SCENE_DEFS})
+
+# ── xgb_boundary_parity — prove C++ scene-boundary inference matches Python ───
+if(SAE_SCENE_XGB)
+    add_executable(xgb_boundary_parity src/tools/xgb_boundary_parity.cpp)
+    target_include_directories(xgb_boundary_parity PRIVATE src ${HDF5_INCLUDE_DIRS})
+    target_link_libraries(xgb_boundary_parity PRIVATE
+        xgboost ${HDF5_CXX_LIBRARIES} ${FFTW3_LIB} ffmpeg_libs)
+
+    # Dumps the C++ feature matrix so training uses the exact inference features.
+    add_executable(scene_features_dump src/tools/scene_features_dump.cpp)
+    target_include_directories(scene_features_dump PRIVATE src ${HDF5_INCLUDE_DIRS})
+    target_link_libraries(scene_features_dump PRIVATE
+        xgboost ${HDF5_CXX_LIBRARIES} ${FFTW3_LIB} ffmpeg_libs)
+endif()
 
 # ── analyze_debug — same binary with debug frame/crop output ─────────────────
 add_executable(scene_analyze_debug src/main.cpp)
-target_link_libraries(scene_analyze_debug PRIVATE sae_gallery ${HDF5_CXX_LIBRARIES})
+target_link_libraries(scene_analyze_debug PRIVATE sae_gallery ${HDF5_CXX_LIBRARIES} ${SAE_SCENE_LIBS})
 target_include_directories(scene_analyze_debug PRIVATE ${HDF5_INCLUDE_DIRS})
-target_compile_definitions(scene_analyze_debug PRIVATE SAE_DEBUG=1)
+target_compile_definitions(scene_analyze_debug PRIVATE SAE_DEBUG=1 ${SAE_SCENE_DEFS})
 
 # ── dump_embeddings — standalone embedding dumper, NO gallery/matcher ─────────
 # Front-half only (decode→detect→align→embed→HDF5) for the optimizer replay corpus

docs/SPEC.md

@@ -128,10 +128,43 @@ Two consequences worth stating:
   depends on timing. The same command run twice can produce different dumps, and
   a golden fixture cannot be built on that.
 
-**Current:** fixed in KPN — node data outputs use `push_blocking`, sentinels
+**Current:** fixed in KPN. Node data outputs *park* on a full channel — the
-remain out-of-band so EOF can always overtake a stalled data path. Verified on
+value is held in a one-slot buffer, the worker is released, and the channel's
-the same clip: 385 of 385 sampled frames written, zero drops, and two
+space callback resubmits the node once the consumer drains. That replaced
-consecutive runs byte-identical where previously they were not.
+`push_blocking`, which slept inside the push and, with one thread per node,
+stopped that node draining its own input. Sentinels remain out-of-band so EOF
+can always overtake a stalled data path. Verified on the same clip: 385 of 385
+sampled frames written, zero drops, and two consecutive runs byte-identical
+where previously they were not.
+
+A later audit found the losslessness was still incomplete in three places, all
+now closed and each pinned by a regression case in the KPN suite:
+
+- **`FilterNode` and `RouterNode`** were the last data paths still using the
+  throwing `push()` with the exception swallowed. A full output discarded the
+  value, and that included the **EOF sentinel**. The decimator passes EOF by
+  predicate but its output is reliably full — the embedder is the slowest node
+  in the chain — so the token was discarded, nothing downstream shut down, and
+  the run had to be killed. This was the wedge.
+- **The sentinel could arrive ahead of a value still queued behind it.** `pop()`
+  observed the ring empty and then took the sentinel; a producer can push a
+  value *and* publish the sentinel inside that window, so a consumer treating
+  EOF as a hard stop loses the tail.
+- **Two firings of one node could overlap**, because the submit gate was
+  released before the firing had finished with the node's state. That breaks the
+  one-slot park itself: a parked value can be overwritten by the other firing,
+  with no drop recorded anywhere.
+
+**New constraint:** a channel carries at most one undelivered sentinel. A second
+offered before the first is taken is refused and reported, never queued and
+never overwritten — two control tokens on one channel means the stream ended
+twice. Single-shot EOF is what everything does today; this becomes live the
+moment a pipeline is reused for a second input.
+
+**Consequence:** a lossless decimator is a backpressure point, not a relief
+valve. The source now throttles to the face branch rather than quietly thinning
+it. That is what this requirement asks for, but it changes the shape of a loaded
+run and has not yet been benchmarked.
 
 It also ran *faster* (29 s → 17 s). A dropped frame has already cost its decode,
 and the overflow exception cost more — so the lossy path was paying for work it

docs/best-model-2026-07.md

@@ -1,10 +1,12 @@
+> **Archived (July 2026).** This report covers the pre-opencv5 framework and the 4-model ArcFace/LVFace bake-off. It is superseded by the current [experiment log](model-bakeoff.md) for the opencv5 build. Kept for provenance; the numbers here are historical.
+
 # Which embedding model is best?
 
 Three ArcFace variants (w600k-R50, R18, w600k-MBF) and LVFace-B (Glint360K,
 455MB) were compared. r50 is excluded from the training/held-out comparison
 below; its gallery has roughly 30% fewer reference images per actor than the
 other three on the identical source photos, which confounds a direct score
-comparison (see [the full experiment log](model-bakeoff.md) for detail). It
+comparison (see [the full experiment log](model-bakeoff-2026-07.md) for detail). It
 remains in the calibration comparison, which does not depend on the gallery
 image count.
 
@@ -63,7 +65,7 @@ than general performance. On training data, the ordering is not as clean:
 
 mbf beats LVFace on Lord of War (77.2% vs 75.6%), the only film in either
 table where LVFace does not score highest. LVFace's training-set macro
-average (75.3%, see [the full experiment log](model-bakeoff.md)) is not a
+average (75.3%, see [the full experiment log](model-bakeoff-2026-07.md)) is not a
 uniform win across every film it contributes to; the held-out result, where
 LVFace wins all 5 films outright, is the stronger claim.
 
@@ -79,7 +81,7 @@ not.
 ![All 12 combos ranked by training-set F1](assets/images/rep4_matrix_f1.png)
 
 Best full-gallery combo per model (all three are `full_exp`), from the
-training matrix in [the full experiment log](model-bakeoff.md):
+training matrix in [the full experiment log](model-bakeoff-2026-07.md):
 
 | model | F1 | P | R | misID |
 |---|---|---|---|---|

docs/gallery-scope-2026-07.md

@@ -1,3 +1,5 @@
+> **Archived (July 2026).** This report covers the pre-opencv5 framework and the 4-model ArcFace/LVFace bake-off. It is superseded by the current [experiment log](model-bakeoff.md) for the opencv5 build. Kept for provenance; the numbers here are historical.
+
 # Whole gallery vs. cast-restricted gallery
 
 Two ways to run the matcher. Full mode scores every detected face against
@@ -8,7 +10,7 @@ top-billed actors) before the matcher runs.
 ## Result
 
 Averaged across the 3 compared models (r50 excluded, see
-[the full experiment log](model-bakeoff.md)) and both expansion settings, on
+[the full experiment log](model-bakeoff-2026-07.md)) and both expansion settings, on
 the 4 training films:
 
 | scope | F1 | P | R | total misID |
@@ -27,7 +29,7 @@ restricted gallery:
 
 ![All combos ranked by training-set F1, filled dots are restricted](assets/images/rep4_matrix_f1.png)
 
-See [the full experiment log](model-bakeoff.md) for the complete table. One
+See [the full experiment log](model-bakeoff-2026-07.md) for the complete table. One
 combo reaches zero true out-of-cast misidentifications,
 `arcface_w600k_mbf_restricted_exp` (F1 76.2%), and it is a restricted one,
 consistent with restriction, not expansion, being what suppresses cross-film
@@ -57,7 +59,7 @@ Building this as a real feature requires:
   option.
 - A decision on the fallback case: what happens to a real, uncredited
   cameo (see the Germar Terrell Gardner and Talia Balsam cases in the
-  [LVFace deep dive](lvface-deep-dive.md#where-lvface-beat-x-ray)) if the
+  [LVFace deep dive](lvface-deep-dive-2026-07.md#where-lvface-beat-x-ray)) if the
   restricted gallery never includes them at all.
 - Regenerating the restricted-gallery cache whenever a title's Jellyfin
   cast list changes.

docs/index.md

@@ -17,62 +17,70 @@ two credited cast members without a visible face are correctly reported
 present but not visible. This matches Amazon X-Ray's own record for this
 second exactly.
 
-Results are not uniform across films. The hardest held-out film scores 46%
+## The headline: learned scene boundaries
-F1. This report documents why: one tunable trade (extinction bridging at
-hard cuts), one structural limit (X-Ray credits people whose faces never
-appear on screen), and a small number of cases where the pipeline is
-correct and X-Ray's ground truth is not. Read
-[how we score against X-Ray](methodology.md) first. X-Ray's ground truth is
-scene-level; the pipeline's output is per-second. That difference shapes
-every finding below.
 
-## Findings
+The current opencv5 build's biggest gain is **flood-fill presence on a
+learned scene-boundary detector**. An actor seen once inside a shot is
+reported for the whole shot — but only if the shot boundaries are good. A
+learned XGBoost boundary detector, scored **leave-one-out** so no film is
+ever measured by a detector that trained on it, lifts per-second X-Ray
+presence F1 across nine films and improves every one of them:
 
-<div class="grid cards" markdown>
+| boundary source for flood-fill | presence F1 |
+| ------------------------------ | ----------: |
+| track-extent (flood off) | 62.6% |
+| flood + grayscale cuts | 64.0% |
+| **flood + learned detector (LOO)** | **74.9%** |
 
--   :material-trophy:{ .lg .middle } **[Which model is best?](best-model.md)**
+![Macro presence F1 by flood-fill boundary source](assets/images/scene_presence_macro.png)
 
-    ---
+The full story — why the old grayscale cut detector broke Scarface, what
+features work, and the per-film breakdown — is on the
+[learned scene-boundary detector](scene-boundary-detector.md) page.
 
-    Calibration curves first, independent of any threshold, then held-out
+## What the numbers mean, and their limits
-    F1 across three models. LVFace-B Glint360K wins both, and wins on every
-    held-out film.
 
--   :material-filter:{ .lg .middle } **[Whole vs. cast-restricted gallery](gallery-scope.md)**
+Results are not uniform across films, and they should not be. X-Ray's ground
+truth is scene-level and credits people whose faces never appear on screen;
+the pipeline's output is per-second and can only name a face it can see.
+That difference is a structural recall ceiling, not a bug. Read
+[how we score against X-Ray](methodology.md) first — it defines F1,
+precision, recall, and misID, and explains the two limits (off-screen cast
+and gallery coverage) that shape every finding.
 
-    ---
+Precision on identified faces is near-perfect: where the pipeline names a
-
+face, it is almost always a name X-Ray also credits to that scene. The
-    Restricting the matcher to a film's credited cast improves F1,
+frames throughout this documentation make the tension visual — **green** =
-    recall, and misID rate at once, but is not a shipped runtime feature
+true positive, **red** = false positive, **orange** = unknown, and a
-    yet.
+**blue** panel lists credited cast present with no visible face.
-
--   :material-account-convert:{ .lg .middle } **[Does pose expansion help?](pose-expansion.md)**
-
-    ---
-
-    A training-set effect that did not reproduce on 5 held-out films once
-    two methodology bugs in the comparison harness were found and fixed.
-
--   :material-magnify-expand:{ .lg .middle } **[Deep dive: LVFace-B Glint360K](lvface-deep-dive.md)**
-
-    ---
-
-    The held-out generalization gap, the two mechanisms behind its errors,
-    and every distinct case where it names someone outside the film's
-    credited cast.
-
-</div>
 
 ## Full experiment log
 
-- **[Full experiment log](model-bakeoff.md)**: the complete log behind the
+- **[Full experiment log (opencv5)](model-bakeoff.md)**: the complete log
-  four pages above, including how replaying against cached embeddings
+  behind the current build — the ten-knob differential-evolution tuning, the
-  inside the same KPN network makes a full model and configuration
+  shipped [`src/config.hpp`](https://REPOLINK/src/config.hpp) defaults and
-  comparison practical, the full results table, and every caveat. This is
+  where each comes from, the replay architecture that makes a nine-film
-  where the shipped [`src/config.hpp`](https://REPOLINK/src/config.hpp)
+  search tractable, and the flood-fill step change.
-  defaults come from.
+- **[Learned scene-boundary detector](scene-boundary-detector.md)**: the
-- **[Service conversion (proposal)](service-conversion.md)**: design
+  features, the model, leave-one-out results, and the two headline films.
-  sketch for a native idle-GPU worker gated on screen lock, not yet built.
+- **[Benchmark — SuperHero](benchmark.md)**: the benchmark harness.
+- **[Service conversion (proposal)](service-conversion.md)**: design sketch
+  for a native idle-GPU worker gated on screen lock, not yet built.
+
+## Archive (July 2026)
+
+The pre-opencv5 four-model ArcFace/LVFace bake-off is kept for provenance.
+Its numbers are historical; the current build supersedes them.
+
+- [Best model (July)](best-model-2026-07.md) — LVFace-B Glint360K wins on
+  calibration and on every held-out film.
+- [Gallery scope (July)](gallery-scope-2026-07.md) — cast-restricted
+  gallery improves F1, recall, and misID at once.
+- [Pose expansion (July)](pose-expansion-2026-07.md) — a training-set
+  effect that did not reproduce held-out.
+- [LVFace deep dive (July)](lvface-deep-dive-2026-07.md) — the
+  generalization gap and every out-of-cast identification.
+- [Full experiment log (July)](model-bakeoff-2026-07.md).
 
 ## Reproducing the benchmarks
 

docs/lvface-deep-dive-2026-07.md

@@ -1,12 +1,14 @@
+> **Archived (July 2026).** This report covers the pre-opencv5 framework and the 4-model ArcFace/LVFace bake-off. It is superseded by the current [experiment log](model-bakeoff.md) for the opencv5 build. Kept for provenance; the numbers here are historical.
+
 # Deep dive: LVFace-B Glint360K
 
-LVFace won the model comparison (see [Which model is best?](best-model.md))
+LVFace won the model comparison (see [Which model is best?](best-model-2026-07.md))
 and is the shipped default embedder. This page reports how it performs in
 detail: a baseline of correct output, the two mechanisms behind its errors,
 and every distinct case where it names someone who is not in the film's
 credited cast.
 
-Read [How we score against X-Ray](methodology.md) first. X-Ray's ground truth
+Read [How we score against X-Ray](methodology-2026-07.md) first. X-Ray's ground truth
 is scene-level, not per-frame. A name marked correct in the Offscreen column
 below is the pipeline correctly reporting scene membership, not a workaround.
 
@@ -61,7 +63,7 @@ on the 5 films the optimizer never saw:
 | macro average | 67.4% | 85.8% | 57.0% | | | | |
 
 The `P` column is misID-weighted (each out-of-film name counts 10x in the
-denominator; see [methodology](methodology.md#precision-recall-and-the-misid-weighting)).
+denominator; see [methodology](methodology-2026-07.md#precision-recall-and-the-misid-weighting)).
 That weighting is why Many Saints reads 54.7% here despite naming mostly real,
 present faces: its raw (unweighted) precision is **78.4%**, and the gap is
 entirely its 974 misIDs paying the 10x penalty. The three zero-misID films
@@ -70,7 +72,7 @@ Lovelace, with 58 misIDs, sits 3pp below its raw 93.3%.
 
 Held-out F1 is 67.4%, against 75.3% on training, an 8pp drop. The spread
 between the best and worst held-out film is 37pp. This is not unique to
-LVFace: [the full experiment log](model-bakeoff.md#held-out-validation-all-3-models)
+LVFace: [the full experiment log](model-bakeoff-2026-07.md#held-out-validation-all-3-models)
 shows mbf and r18 with the same shape of spread on the same films, at a
 uniformly lower level. Two mechanisms explain the spread. Both are shown
 below with frame-level evidence.
@@ -174,10 +176,10 @@ ground-truth gap, not a model error.
 Archie Yates, t=2521s, 78% confidence. A real detected face, a genuine
 lookalike confusion.
 
-![Zooey Deschanel, third out-of-cast name in Many Saints](assets/images/many_saints_fpi_deschanel.jpg)
+Zooey Deschanel, t=2819s, 99% confidence — a high-confidence lookalike
-
+confusion in the July pipeline. **The current opencv5 pipeline no longer makes
-Zooey Deschanel, t=2819s, 99% confidence. A real detected face at a dinner
+this identification**; the tighter tracker/registry and re-tuned matching removed
-table, high-confidence lookalike confusion.
+it, so there is no annotated frame for it here.
 
 ![Talia Balsam, fourth out-of-cast name in Many Saints](assets/images/many_saints_fpi_balsam.jpg)
 

docs/methodology-2026-07.md

@@ -0,0 +1,136 @@
+> **Archived (July 2026).** This report covers the pre-opencv5 framework and the 4-model ArcFace/LVFace bake-off. It is superseded by the current [experiment log](model-bakeoff.md) for the opencv5 build. Kept for provenance; the numbers here are historical.
+
+# How we score against X-Ray
+
+Every number in this report, every F1 and misID count, comes from one
+comparison. The comparison has a mismatch at its core that shapes nearly
+every finding in this report: the ground truth is scene-level, the
+pipeline's output is per-second, and the two do not mean the same thing.
+This page documents that comparison once, so the findings pages can rely on
+it without re-explaining it.
+
+## What Amazon X-Ray records
+
+X-Ray ships three tables per film: `scenes.csv` (a list of `[start, end]`
+timespans), `people_in_scenes.csv` (which actors are credited in each
+scene), and `people.csv` (actor identities). There is no per-frame or
+per-second annotation anywhere in X-Ray. A scene might run 45 seconds, and
+X-Ray records one cast list for the entire span, not "on screen from
+second 12 to second 30."
+
+To compare this against per-second predictions, `second_score.py` expands
+every scene into per-second ground truth by copying the whole scene's cast
+list onto every second inside it:
+
+```python
+for sn, (t0, t1) in spans.items():
+    cast = scene_cast.get(sn, [])
+    for t in range(int(t0), int(t1)):
+        timeline[t] = cast
+```
+
+That is the entire mechanism. If X-Ray credits five actors to a 30-second
+scene, all five count as ground truth present for all 30 seconds, including
+seconds where only one of them is on screen. This is not a simplification
+introduced by the pipeline; it is the only reading of X-Ray's data that is
+possible, because X-Ray itself does not record anything finer-grained.
+
+## Why an offscreen name can be scored correct
+
+A name listed under Offscreen with a correct (green) label is not the
+pipeline guessing or padding its score. It is the pipeline correctly
+answering the question X-Ray actually asks: is this actor part of this
+scene. It answers that question using a presence window (`[start, end]`,
+held open across cuts by `anneal_sec` and `extinction_sec`), which matches
+X-Ray's scene-level semantics more closely than a raw per-frame detection
+would.
+
+A system that only reported "this actor is visible in this exact frame"
+would score worse against X-Ray's scene-level ground truth, producing a
+false negative every time the camera cuts away from a character who is
+still present in the scene. Not because it is wrong about the world, but
+because it would be answering a stricter, different question than the one
+X-Ray's data supports. The presence-window design exists specifically to
+answer X-Ray's actual question.
+
+## What this resolves and what it does not
+
+This resolves the semantic mismatch between a scene and an instant. It does
+not resolve two other limitations, both discussed in the
+[LVFace deep dive](lvface-deep-dive-2026-07.md).
+
+**The face-vs-presence ceiling.** X-Ray credits scene membership regardless
+of whether a face is ever visible: background crew, characters shot from
+behind, voice-only presence. No amount of bridging recovers a face that
+never appears on screen. This is a hard ceiling on recall, not a defect.
+
+**Extinction bridging can overshoot.** The same presence-window mechanism
+that correctly answers "still in this scene" during a normal cut can also
+bridge across a scene boundary it has no way to detect. A hard cut into a
+different scene with no faces, such as closing credits, carries the
+previous scene's identities forward until the window expires. This is the
+mechanism behind Downton Abbey's recall collapse, documented in the deep
+dive.
+
+## Precision, recall, and the misID weighting
+
+Per sampled second `t`:
+
+**TPI** (true positive instances): actors both X-Ray and the pipeline agree
+are present.
+
+**FPI** (false positive instances): actors the pipeline reports that are
+not in X-Ray's cast for this second. Split into two categories:
+
+- **FPI_incast**: the actor is in the film's cast, just not credited to
+  this particular scene. A timing or boundary slip.
+- **FPI_misid**: the actor is not in the film's cast at all. A genuine
+  wrong-identity error, weighted 10x in the precision objective, because
+  naming someone who is not even in the film is a categorically worse
+  error than a few seconds of scene-boundary slop.
+
+!!! note "Every headline `P` and `F1` is misID-weighted"
+
+    The precision reported throughout this report, and therefore the F1
+    derived from it, puts each `FPI_misid` into the denominator **10 times**
+    (`precision = TPI / (TPI + FPI_incast + 10·FPI_misid)`,
+    [`second_score.py`](https://REPOLINK/scripts/optimizer/second_score.py)).
+    This is deliberate: the whole point is to punish naming an out-of-film
+    actor far harder than a scene-boundary slip. But it means the `P` column
+    is not raw precision, and a misID-heavy film's `P` is depressed
+    super-linearly. `second_score.py` also emits an unweighted `precision_raw`
+    (always ≥ the weighted `P`); where the gap matters, The Many Saints of
+    Newark, weighted `P` 54.7% vs. raw 78.4%, the [LVFace deep dive](lvface-deep-dive-2026-07.md)
+    reports both. When comparing `P` across films, remember you are comparing a
+    quantity that penalizes misIDs, not just a hit rate.
+
+**FN** (false negatives): actors X-Ray lists that the pipeline never
+reports, counted only for actors who have a gallery reference embedding.
+Across the 9-film benchmark, coverage of X-Ray's credited cast ranges from
+20% to 79% by film (see
+[the full experiment log](model-bakeoff-2026-07.md#gallery-coverage-per-film)); an
+actor with no reference photo can never be recognized regardless of model
+quality, and counting them as a miss would penalize gallery coverage, not
+recognition accuracy.
+
+Two further numbers are reported alongside F1:
+
+**agreement_rate**: mean per-second Jaccard overlap
+(`|Pred ∩ GT| / |Pred ∪ GT|`), partial credit. Naming 2 of 3 present actors
+scores 2/3, not 0.
+
+**exact_match_rate**: the fraction of sampled seconds where the pipeline's
+named set exactly equals X-Ray's, no partial credit. Far harsher, and
+dominated by recall, since any single missed actor zeroes that second.
+
+## Reproduce
+
+```bash
+python3 scripts/optimizer/second_score.py \
+    --pred pred.json --xray experiments/xray/.../<xray_dir> \
+    --gallery experiments/galleries/gallery_LVFace-B_Glint360K.h5
+```
+
+See also [the full experiment log](model-bakeoff-2026-07.md) for how `pred.json` is
+produced, and the [LVFace deep dive](lvface-deep-dive-2026-07.md) for what these
+mechanisms look like frame by frame.

docs/methodology.md

@@ -1,11 +1,9 @@
 # How we score against X-Ray
 
-Every number in this report, every F1 and misID count, comes from one
+Every number in this report comes from one comparison, and that comparison
-comparison. The comparison has a mismatch at its core that shapes nearly
+has a mismatch at its core: the ground truth is scene-level, the pipeline's
-every finding in this report: the ground truth is scene-level, the
+output is per-second, and the two do not mean the same thing. This page
-pipeline's output is per-second, and the two do not mean the same thing.
+documents the comparison once so the findings can rely on it.
-This page documents that comparison once, so the findings pages can rely on
-it without re-explaining it.
 
 ## What Amazon X-Ray records
 
@@ -13,12 +11,12 @@ X-Ray ships three tables per film: `scenes.csv` (a list of `[start, end]`
 timespans), `people_in_scenes.csv` (which actors are credited in each
 scene), and `people.csv` (actor identities). There is no per-frame or
 per-second annotation anywhere in X-Ray. A scene might run 45 seconds, and
-X-Ray records one cast list for the entire span, not "on screen from
+X-Ray records one cast list for the entire span, not "on screen from second
-second 12 to second 30."
+12 to second 30."
 
-To compare this against per-second predictions, `second_score.py` expands
+To compare against per-second predictions, `second_score.py` expands every
-every scene into per-second ground truth by copying the whole scene's cast
+scene into per-second ground truth by copying the whole scene's cast list
-list onto every second inside it:
+onto every second inside it:
 
 ```python
 for sn, (t0, t1) in spans.items():
@@ -27,48 +25,46 @@ for sn, (t0, t1) in spans.items():
         timeline[t] = cast
 ```
 
-That is the entire mechanism. If X-Ray credits five actors to a 30-second
+If X-Ray credits five actors to a 30-second scene, all five count as ground
-scene, all five count as ground truth present for all 30 seconds, including
+truth present for all 30 seconds, including seconds where only one is on
-seconds where only one of them is on screen. This is not a simplification
+screen. This is not a simplification the pipeline introduces; it is the only
-introduced by the pipeline; it is the only reading of X-Ray's data that is
+reading X-Ray's data supports, because X-Ray records nothing finer.
-possible, because X-Ray itself does not record anything finer-grained.
 
-## Why an offscreen name can be scored correct
+## How the pipeline reports presence
 
-A name listed under Offscreen with a correct (green) label is not the
+A presence claim is one actor owning one time window. How that window is
-pipeline guessing or padding its score. It is the pipeline correctly
+derived is a tunable choice — a knob the optimizer weighs — with two modes:
-answering the question X-Ray actually asks: is this actor part of this
-scene. It answers that question using a presence window (`[start, end]`,
-held open across cuts by `anneal_sec` and `extinction_sec`), which matches
-X-Ray's scene-level semantics more closely than a raw per-frame detection
-would.
 
-A system that only reported "this actor is visible in this exact frame"
+- **`track_extent` (default).** A claim is exactly `[first_seen, last_seen]`
-would score worse against X-Ray's scene-level ground truth, producing a
+  of a track the actor owned (AR-012), ending at the last sighting and never
-false negative every time the camera cuts away from a character who is
+  after (AR-013). There is no keep-alive: the withdrawn `anneal_sec` and the
-still present in the scene. Not because it is wrong about the world, but
+  scene-tracker `extinction_sec` — which the July report's windows were held
-because it would be answering a stricter, different question than the one
+  open by — are **gone**. A track that survives its own gaps needs no bridge;
-X-Ray's data supports. The presence-window design exists specifically to
+  a gap after the final sighting is never claimed.
-answer X-Ray's actual question.
+- **`flood`.** Each claim is snapped to the shot it sits in, so an actor seen
+  once anywhere in a shot is reported for the whole shot
+  `[prev_boundary, next_boundary]`. Boundaries come from TransNetV2 shot
+  detection when available, otherwise from the always-on histogram cut
+  detector (`is_cut`). This trades precision for recall against X-Ray's
+  scene-level granularity, and the optimizer decides per run whether it pays.
 
-## What this resolves and what it does not
+Do not confuse the surviving `track_extinction_sec` with the withdrawn
+scene `extinction_sec`: the former bounds how long a lost track stays
+available for **re-association** (a tracking question), and never extends a
+presence claim.
 
-This resolves the semantic mismatch between a scene and an instant. It does
+## The two limits this does not resolve
-not resolve two other limitations, both discussed in the
-[LVFace deep dive](lvface-deep-dive.md).
 
 **The face-vs-presence ceiling.** X-Ray credits scene membership regardless
 of whether a face is ever visible: background crew, characters shot from
-behind, voice-only presence. No amount of bridging recovers a face that
+behind, voice-only presence. No face pipeline can recover a face that never
-never appears on screen. This is a hard ceiling on recall, not a defect.
+appears, so recall against X-Ray is a structural ceiling, not a defect.
 
-**Extinction bridging can overshoot.** The same presence-window mechanism
+**Flood-fill can overshoot.** Snapping to a shot correctly answers "still in
-that correctly answers "still in this scene" during a normal cut can also
+this scene" through an intra-scene cut, but a shot boundary is not a scene
-bridge across a scene boundary it has no way to detect. A hard cut into a
+boundary: on a film with sparse cuts, flood-fill can carry an actor across a
-different scene with no faces, such as closing credits, carries the
+long "shot" they only briefly appeared in. This is why flood-fill is a knob,
-previous scene's identities forward until the window expires. This is the
+not a default — its value depends on the film's cut density.
-mechanism behind Downton Abbey's recall collapse, documented in the deep
-dive.
 
 ## Precision, recall, and the misID weighting
 
@@ -77,49 +73,46 @@ Per sampled second `t`:
 **TPI** (true positive instances): actors both X-Ray and the pipeline agree
 are present.
 
-**FPI** (false positive instances): actors the pipeline reports that are
+**FPI** (false positive instances): actors the pipeline reports that are not
-not in X-Ray's cast for this second. Split into two categories:
+in X-Ray's cast for this second, split into:
 
-- **FPI_incast**: the actor is in the film's cast, just not credited to
+- **FPI_incast**: the actor is in the film's cast, just not credited to this
-  this particular scene. A timing or boundary slip.
+  scene. A timing or boundary slip.
 - **FPI_misid**: the actor is not in the film's cast at all. A genuine
-  wrong-identity error, weighted 10x in the precision objective, because
+  wrong-identity error, weighted **10×** in the precision objective, because
-  naming someone who is not even in the film is a categorically worse
+  naming someone not even in the film is categorically worse than a few
-  error than a few seconds of scene-boundary slop.
+  seconds of scene-boundary slop.
 
 !!! note "Every headline `P` and `F1` is misID-weighted"
 
-    The precision reported throughout this report, and therefore the F1
+    Precision puts each `FPI_misid` into the denominator 10 times
-    derived from it, puts each `FPI_misid` into the denominator **10 times**
     (`precision = TPI / (TPI + FPI_incast + 10·FPI_misid)`,
     [`second_score.py`](https://REPOLINK/scripts/optimizer/second_score.py)).
-    This is deliberate: the whole point is to punish naming an out-of-film
+    This deliberately punishes naming an out-of-film actor far harder than a
-    actor far harder than a scene-boundary slip. But it means the `P` column
+    boundary slip, so the `P` column is not raw precision and a misID-heavy
-    is not raw precision, and a misID-heavy film's `P` is depressed
+    film's `P` is depressed super-linearly.
-    super-linearly. `second_score.py` also emits an unweighted `precision_raw`
-    (always ≥ the weighted `P`); where the gap matters, The Many Saints of
-    Newark, weighted `P` 54.7% vs. raw 78.4%, the [LVFace deep dive](lvface-deep-dive.md)
-    reports both. When comparing `P` across films, remember you are comparing a
-    quantity that penalizes misIDs, not just a hit rate.
 
-**FN** (false negatives): actors X-Ray lists that the pipeline never
+**FN** (false negatives): actors X-Ray lists that the pipeline never reports,
-reports, counted only for actors who have a gallery reference embedding.
+counted **only** for actors who have a gallery reference embedding. An actor
-Across the 9-film benchmark, coverage of X-Ray's credited cast ranges from
+with no reference photo can never be recognized, and counting them as a miss
-20% to 79% by film (see
+would measure gallery coverage, not recognition accuracy.
-[the full experiment log](model-bakeoff.md#gallery-coverage-per-film)); an
-actor with no reference photo can never be recognized regardless of model
-quality, and counting them as a miss would penalize gallery coverage, not
-recognition accuracy.
 
-Two further numbers are reported alongside F1:
+Two further numbers accompany F1:
 
 **agreement_rate**: mean per-second Jaccard overlap
-(`|Pred ∩ GT| / |Pred ∪ GT|`), partial credit. Naming 2 of 3 present actors
+(`|Pred ∩ GT| / |Pred ∪ GT|`) — partial credit, so naming 2 of 3 present
-scores 2/3, not 0.
+actors scores 2/3, not 0.
 
-**exact_match_rate**: the fraction of sampled seconds where the pipeline's
+**exact_match_rate**: the fraction of seconds where the pipeline's named set
-named set exactly equals X-Ray's, no partial credit. Far harsher, and
+exactly equals X-Ray's — no partial credit, dominated by recall.
-dominated by recall, since any single missed actor zeroes that second.
+
+## The benchmark set
+
+Unlike the July report — which trained on a 3-film subset and validated on
+held-out films to keep evaluations fast — this run scores **all 9 films on
+every evaluation**. The registry one-clock fix and uncapped dumps made
+full-set replay affordable, so the reported optimum is tuned against the
+complete set rather than a training subset.
 
 ## Reproduce
 
@@ -129,6 +122,5 @@ python3 scripts/optimizer/second_score.py \
     --gallery experiments/galleries/gallery_LVFace-B_Glint360K.h5
 ```
 
-See also [the full experiment log](model-bakeoff.md) for how `pred.json` is
+See the [full experiment log](model-bakeoff.md) for how `pred.json` is
-produced, and the [LVFace deep dive](lvface-deep-dive.md) for what these
+produced and where the shipped `src/config.hpp` defaults come from.
-mechanisms look like frame by frame.

docs/model-bakeoff-2026-07.md

@@ -0,0 +1,348 @@
+> **Archived (July 2026).** This report covers the pre-opencv5 framework and the 4-model ArcFace/LVFace bake-off. It is superseded by the current [experiment log](model-bakeoff.md) for the opencv5 build. Kept for provenance; the numbers here are historical.
+
+# Full experiment log
+
+This page reports how the pipeline performs across three questions: which
+embedding model is best, whether restricting the gallery to a film's
+credited cast helps, and whether promoting confidently identified poses into
+a per-film gallery annex helps. It also documents the replay architecture
+that made testing all three questions in one pass practical, and every
+caveat needed to trust the numbers.
+
+Read [How we score against X-Ray](methodology-2026-07.md) first for what F1,
+precision, recall, and misID mean in this report. All numbers below use the
+per-second metric
+([`scripts/optimizer/second_score.py`](https://REPOLINK/scripts/optimizer/second_score.py)).
+
+r50 (ArcFace w600k-R50) is excluded from the detailed comparison below. Its
+gallery was built with roughly 30% fewer reference images per actor than the
+other three models on the identical source photos (10808 vs 15055 total
+embeddings across the same 2418 actors), which confounds any direct
+comparison of its scores against the others. It remains in the
+[calibration curve comparison](best-model-2026-07.md#first-signal-calibration-curves),
+which does not depend on the training benchmark.
+
+## Why replay makes this affordable
+
+Decoding video and running face detection, alignment, and embedding is the
+expensive part of this pipeline. Everything downstream of that (tracking,
+identity matching, scene aggregation) is cheap. KPN++'s node/network
+structure means those two stages are separate components connected by
+typed channels, so the expensive stage can run once per film, cache its
+output, and the cheap stage can be re-run against that cache as many times
+as needed with different Config values.
+
+`scene_analyze --dump-embeddings out.h5` runs the expensive half once per
+film and writes per-frame face detections and embeddings to HDF5
+([`scripts/optimizer/SCHEMA.md`](https://REPOLINK/scripts/optimizer/SCHEMA.md)).
+[`scripts/optimizer/replay.py`](https://REPOLINK/scripts/optimizer/replay.py)
+then re-assembles the real C++ `face_tracker`, `identity_matcher`, and
+`scene_tracker` nodes into a Python-driven KPN network and replays a
+film's cached embeddings through them, varying `prob_threshold`,
+`anneal_sec`, `extinction_sec`, and `expand_gallery` freely. No GPU
+inference and no video decode happen during a replay; each one completes
+in seconds. This is what makes a 512-evaluation differential-evolution
+search per model, per gallery mode, per expansion setting, tractable, and
+what made the full held-out validation across three models in this report
+possible in one session rather than requiring three full re-encodes of the
+benchmark set.
+
+`optimize.py` runs `differential_evolution` over this replay function as its
+objective, with DE-level parallelism (multiple candidate configs evaluated
+concurrently, each spawning its own replay subprocesses) on top of it. The
+practical ceiling on this machine's GPU was 8 concurrent replay processes;
+9 silently degraded every score to 0.0% (well-formed output, wrong numbers,
+not a crash), so `optimize.py` was run at `REPLAY_WORKERS=4 DE_WORKERS=2`.
+
+## Search space
+
+`popsize=10, maxiter=15` per combo (3 parameters, up to 512 evaluations,
+usually stopping earlier on DE's convergence tolerance).
+`anneal_sec`/`extinction_sec` bounds were widened from 1-30/1-15 to 1-60/1-60
+partway through the sweep. r50's 4 combos finished before the widening and
+used the old, narrower bounds; this is one more reason r50 is excluded from
+direct comparison here.
+
+## Training films and held-out films
+
+9 films have dumped embeddings across all 4 models. 4 were used for
+optimization:
+
+- Café Society (62-cast)
+- Lord of War (64-cast)
+- Scarface (67-cast)
+- Sound of Metal (14-cast)
+
+5 were held out, never seen by any optimizer run:
+
+- Benny & Joon
+- Downton Abbey: A New Era
+- Lovelace
+- The Many Saints of Newark
+- Valerian and the City of a Thousand Planets
+
+## Gallery coverage per film
+
+The gallery has reference embeddings for 2418 actors, but coverage of any
+given film's credited cast varies widely. This was previously reported as
+one flat number (67% of X-Ray cast lacking a reference embedding, averaged
+across the whole benchmark); the per-film breakdown is:
+
+| film | cast credited | in gallery | coverage |
+|---|---|---|---|
+| Lord of War | 64 | 13 | 20.3% |
+| Scarface | 67 | 15 | 22.4% |
+| The Many Saints of Newark | 48 | 13 | 27.1% |
+| Café Society | 62 | 17 | 27.4% |
+| Lovelace | 42 | 15 | 35.7% |
+| Valerian and the City of a Thousand Planets | 36 | 13 | 36.1% |
+| Benny & Joon | 23 | 12 | 52.2% |
+| Downton Abbey: A New Era | 36 | 22 | 61.1% |
+| Sound of Metal | 14 | 11 | 78.6% |
+
+Two training films (Lord of War, Scarface) have the worst coverage in the
+set, 20-22%. Their training-set F1 numbers below are partly capped by
+missing references, not purely by model quality. Downton Abbey has 61%
+coverage, the second-best in the benchmark, yet the worst held-out recall
+of any film (39.4%, LVFace). Its recall problem is not primarily a coverage
+problem; it is the extinction-bridging failure documented in the
+[LVFace deep dive](lvface-deep-dive-2026-07.md#mechanism-1-extinction-bridging).
+Reproduce with `scripts/docs/gallery_coverage_per_film.py`.
+
+## Training results, 3 models × 2 gallery modes × 2 expansion settings
+
+Ranked by F1. misid = FPI_misid, the count of true wrong-actor
+identifications (naming someone not in the film's cast at all), distinct
+from FPI, which also includes in-cast timing slips.
+
+Each combo's row is its best **full-coverage** evaluation: the highest-F1 DE
+evaluation in which all 4 training films replayed without a timeout (see
+[Dropped-film scoring](#a-scoring-bug-worth-recording-dropped-film-evaluations)
+below for why this qualifier is load-bearing and not the same as `argmax F1`
+over the raw sweep).
+
+| combo | F1 | P | R | TPI | FPI | misid | FN |
+|---|---|---|---|---|---|---|---|
+| LVFace-B_Glint360K_restricted_exp | 78.3% | 91.0% | 68.9% | 42830 | 3782 | 60 | 19492 |
+| LVFace-B_Glint360K_restricted_noexp | 76.7% | 91.5% | 66.2% | 41149 | 3400 | 59 | 21173 |
+| arcface_w600k_mbf_restricted_exp | 76.2% | 90.0% | 66.2% | 64328 | 7480 | 0 | 33234 |
+| arcface_r18_restricted_exp | 75.5% | 87.6% | 66.5% | 41399 | 5666 | 60 | 20923 |
+| LVFace-B_Glint360K_full_exp | 75.3% | 89.7% | 65.4% | 47757 | 3407 | 232 | 26966 |
+| arcface_w600k_mbf_restricted_noexp | 75.0% | 91.1% | 63.9% | 39752 | 3465 | 60 | 22570 |
+| arcface_r18_restricted_noexp | 73.5% | 91.3% | 61.7% | 38299 | 3220 | 60 | 24023 |
+| LVFace-B_Glint360K_full_noexp | 72.3% | 88.3% | 61.8% | 40363 | 3503 | 244 | 25850 |
+| arcface_w600k_mbf_full_exp | 72.0% | 87.7% | 61.4% | 39875 | 3729 | 240 | 26338 |
+| arcface_w600k_mbf_full_noexp | 71.0% | 93.2% | 57.9% | 41699 | 2472 | 56 | 33024 |
+| arcface_r18_full_exp | 69.1% | 87.6% | 57.7% | 37342 | 3119 | 242 | 28871 |
+| arcface_r18_full_noexp | 66.6% | 91.3% | 53.1% | 34314 | 2362 | 107 | 31899 |
+
+![All combos ranked by training-set F1](assets/images/rep4_matrix_f1.png)
+
+The two clearest patterns: every model's best-scoring combo uses the
+restricted gallery, and LVFace leads within both gallery modes. `full_exp`
+(the shipped combination) is the best-scoring option that uses only
+features the running application currently supports; restriction is not
+wired into the application yet (see
+[Whole vs. cast-restricted gallery](gallery-scope-2026-07.md)).
+
+### A scoring bug worth recording: dropped-film evaluations
+
+The numbers above are corrected ones. The raw `rep4_best_*.json` files, and an
+earlier version of this table, reported a different `arcface_w600k_mbf_full_noexp`
+row: **74.2% F1 at TPI 12645**, a third the TPI of every sibling combo. That was
+not a better config; it was an artifact of how the optimizer aggregates.
+
+`optimize.py` builds each candidate's score from only the films whose replay
+subprocess returned (`per_film = [m for m in ex.map(_one, films) if m is not
+None]`), then **averages** F1/precision/recall and **sums** TPI/FPI/misID over
+just those survivors. When a film's replay times out (the sweep ran near the
+8-process concurrency ceiling, so this happened intermittently), that film
+silently drops from both. A candidate whose hardest film timed out is therefore
+scored on an easier subset, and differential evolution, maximizing that score,
+will happily converge onto exactly such a candidate. For `mbf_full_noexp` the
+reported winner was one of 7 evaluations (out of 512) whose TPI had collapsed to
+a partial-film subset; its median-coverage evaluations sit around 51686 TPI.
+
+The fix here was to re-derive each combo's best row from its DE trajectory
+(`experiments/trajectories/rep4_*.jsonl`), keeping only evaluations within 30% of
+that combo's median TPI (full 4-film coverage) before taking the best F1. This
+needs no re-running, the honest best configuration was already in the sweep,
+just not the one `argmax F1` selected. Three combos moved: `mbf_full_noexp`
+74.2% → **71.0%**, `LVFace_full_noexp` 72.4% → **72.3%** (and its misID, 0 → 244,
+was itself a dropped-film artifact), `mbf_restricted_exp` 76.5% → **76.2%**. The
+shipped LVFace `full_exp` winner was unaffected, its reported evaluation already
+had full coverage (TPI 47757 ≈ median). `experiment_charts.py` applies the same
+`clean_best` filter, so every figure on this page matches the corrected table.
+The underlying `optimize.py` aggregation is also being fixed so a dropped-film
+evaluation can never be selected as a winner again.
+
+### Per-film training breakdown
+
+The 75.3% LVFace training figure is a macro average across 4 films, not a
+uniform result:
+
+| film | LVFace F1 | mbf F1 | r18 F1 | best model |
+|---|---|---|---|---|
+| Café Society | 68.1% | 62.2% | 60.1% | LVFace |
+| Lord of War | 75.6% | 77.2% | 75.6% | mbf |
+| Scarface | 71.5% | 68.6% | 64.1% | LVFace |
+| Sound of Metal | 78.8% | 76.5% | 71.6% | LVFace |
+
+LVFace does not win every training film. mbf scores higher on Lord of War
+(77.2% vs 75.6%). LVFace's own training-film range is 68.1% to 78.8%, a
+10.7pp spread, smaller than the 37pp spread seen on held-out films but real.
+Reproduce with `scripts/docs/run_holdout_all_models.py --films training`.
+
+## Held-out validation, all 3 models
+
+The training matrix above is training-set fit. Each model's own tuned
+`full_exp` config was replayed against the 5 held-out films, scored the
+same way:
+
+| film | LVFace F1 | mbf F1 | r18 F1 |
+|---|---|---|---|
+| Benny & Joon | 83.0% | 78.5% | 77.1% |
+| Lovelace | 77.5% | 73.7% | 72.2% |
+| Valerian and the City of a Thousand Planets | 74.1% | 70.2% | 71.0% |
+| Downton Abbey: A New Era | 56.2% | 55.0% | 53.0% |
+| The Many Saints of Newark | 46.3% | 44.5% | 42.1% |
+| **macro average** | **67.4%** | **64.4%** | **63.1%** |
+
+LVFace scores highest on every one of the 5 held-out films; the ranking
+never flips. Total misIDs across the 5 films: LVFace 1032, mbf 2197, r18
+1224. LVFace has less than half mbf's misID count while also scoring
+higher on every film. This directly confirms the model choice out of
+sample; it is not inferred from the training numbers alone. See the
+[LVFace deep dive](lvface-deep-dive-2026-07.md) for frame-level detail on where and
+why LVFace still fails on the two worst films. Reproduce with
+`scripts/docs/run_holdout_all_models.py`.
+
+## Two effects in isolation: gallery scope and pose expansion
+
+Averaging across the 3 compared models (r50 excluded) isolates each variable
+from model choice.
+
+**Gallery scope**, averaged over both expansion settings and all 3 models
+(6 evaluations per row):
+
+| scope | F1 | P | R | total misID |
+|---|---|---|---|---|
+| full | 71.1% | 89.6% | 59.6% | 1121 |
+| restricted | 75.9% | 90.4% | 65.6% | 299 |
+
+Restriction improves every metric at once. This is not a precision/recall
+trade: +4.8pp F1, +6.0pp recall, and roughly a quarter the misIDs. Fewer
+candidates in the matcher's search space means fewer opportunities for a
+lookalike false match, and the recall gain shows this does not cost real
+detections. Restriction is currently an offline optimizer technique, not a
+runtime feature of the application; see
+[Whole vs. cast-restricted gallery](gallery-scope-2026-07.md) for what building it
+into the application would require.
+
+**Pose expansion** (promoting a confidently identified track's novel-pose
+views into a per-film gallery annex,
+[`src/gallery/track_gallery.hpp`](https://REPOLINK/src/gallery/track_gallery.hpp)):
+
+| scope | expansion | F1 | R | misID |
+|---|---|---|---|---|
+| full | off | 70.0% | 57.6% | 407 |
+| full | on | 72.1% | 61.5% | 714 |
+| restricted | off | 75.1% | 63.9% | 179 |
+| restricted | on | 76.7% | 67.2% | 120 |
+
+In restricted mode, expansion is a clean win: +1.6pp F1, +3.3pp recall,
+misID drops. The annex only competes against the film's own roughly 15-actor
+cast, so a new pose of a known actor is unlikely to be confused with someone
+else. In full mode, expansion buys +2.1pp F1 and +3.9pp recall but at a real
+cost: misID rises from 407 to 714 as the same new-pose view now competes
+against the full 2418-actor gallery, where a confidently learned pose is more
+likely to match the wrong person. On the full gallery it is a recall-vs-misID
+trade, not a free gain. This training-set effect
+did not reproduce on held-out data; see
+[Does pose expansion help?](pose-expansion-2026-07.md) for the full held-out test
+and the two methodology bugs caught while checking it.
+
+## Calibration curves
+
+Each gallery carries a fitted Platt sigmoid `P(match | sim) = σ(a·sim + b)`,
+stored directly in the gallery HDF5
+([`src/gallery/gallery_calibration.hpp`](https://REPOLINK/src/gallery/gallery_calibration.hpp)).
+This measures discriminative power independent of whatever
+`prob_threshold` a given run used:
+
+![Calibrated P(match|similarity) for all four models](assets/images/calibration_curves.png)
+
+LVFace has the steepest curve (`a=17.7` vs 15.3-16.2 for the ArcFace
+variants) and the lowest P=0.5 decision boundary (similarity 0.23 vs
+0.27-0.31), separating same-actor from different-actor pairs more
+confidently at a lower similarity than any ArcFace variant tested,
+including r50. Generated by
+[`scripts/docs/calibration_chart.py`](https://REPOLINK/scripts/docs/calibration_chart.py).
+
+## Extinction and anneal window search
+
+Every one of the 512 DE evaluations for the winning LVFace `full_exp`
+combo, plotted over the `prob_threshold` × `extinction_sec` plane:
+
+![DE search landscape: 512 evaluations over prob_threshold × extinction_sec](assets/images/de_search_landscape.png)
+
+Nearly everything scoring well sits at `extinction_sec` above 50, across a
+wide range of thresholds. Short extinction windows are uniformly weaker:
+under a strict threshold, there is no good configuration in that region of
+the search space. The optimizer converged with `anneal_sec=59.2,
+extinction_sec=59.2`, about 99% of the widened 60s bound, which raises an
+open question not resolved in this round: does performance keep improving
+past 60s, or does it plateau there. Not chased further this pass.
+
+## Caveats
+
+- r50's 4 combos used the older, narrower search bounds (1-30/1-15 instead
+  of 1-60/1-60) and are further confounded by its thinner gallery. Excluded
+  from all comparisons above except calibration.
+- The shipped defaults use `full_exp` (75.3% training F1), not the
+  higher-scoring `restricted_exp` (78.3%), because cast restriction is not
+  a runtime feature of the application yet.
+- `expand_gallery` is mode-dependent, not a free win. Averaged across models
+  on the full gallery it trades misIDs for recall (see the pose-expansion
+  table). For LVFace specifically, though, `full_exp` beats `full_noexp` on
+  every axis at once (F1 75.3 vs 72.3, precision 89.7 vs 88.3, recall 65.4 vs
+  61.8, misID 232 vs 244), so the shipped `full_exp` is a clean choice for
+  this model, not an F1-vs-safety trade. (An earlier version of this page
+  reported `full_noexp` at 72.4% with zero misIDs and higher precision, which
+  made it look like the safer option; that was the dropped-film artifact
+  described above, not a real property of the config.)
+- Switching the default model is an operational change: any gallery built
+  from a different model's embeddings must be rebuilt before the new
+  default takes effect.
+
+## Reproduce
+
+```bash
+# 4-film training matrix, all 4 models × 2 gallery modes × 2 expansion settings
+bash experiments/run_rep4_subprocess.sh
+
+# single combo
+SAE_EXPAND=1 REPLAY_WORKERS=4 DE_WORKERS=2 python3 scripts/optimizer/optimize.py \
+    --manifest experiments/manifests/rep4_LVFace-B_Glint360K_full.json \
+    --gallery experiments/galleries/gallery_LVFace-B_Glint360K.h5 \
+    --params prob_threshold:0.5:0.999 anneal_sec:1:60 extinction_sec:1:60 \
+    --popsize 10 --maxiter 15 --trajectory traj.jsonl --out best.json
+
+# held-out validation, all 3 models, 5 films
+python3 scripts/docs/run_holdout_all_models.py --out docs_data/holdout_all_models.json
+
+# per-film training breakdown, all 3 models, 4 films
+python3 scripts/docs/run_holdout_all_models.py --films training --out docs_data/training_per_film.json
+
+# gallery coverage per film
+python3 scripts/docs/gallery_coverage_per_film.py --out docs_data/gallery_coverage_per_film.json
+
+# regenerate this page's charts from experiments/ artifacts
+python3 scripts/docs/experiment_charts.py --out-dir docs/assets/images
+
+# one frame per distinct out-of-cast name across all 9 films (used in the deep dive)
+python3 scripts/docs/first_fpi_frames.py
+```
+
+See also the session log
+[`experiments/SESSION_STATE.md`](https://REPOLINK/experiments/SESSION_STATE.md).

docs/model-bakeoff.md

@@ -1,346 +1,198 @@
-# Full experiment log
+# Full experiment log (opencv5)
 
-This page reports how the pipeline performs across three questions: which
+This is the complete log behind the current opencv5 build: how the pipeline is
-embedding model is best, whether restricting the gallery to a film's
+tuned, what the shipped configuration is and where every number in it comes from,
-credited cast helps, and whether promoting confidently identified poses into
+and how the learned scene-boundary detector took per-second actor-presence F1 from
-a per-film gallery annex helps. It also documents the replay architecture
+the low-60s to **74.9%** across the nine-film Amazon X-Ray benchmark — under honest
-that made testing all three questions in one pass practical, and every
+leave-one-out.
-caveat needed to trust the numbers.
 
-Read [How we score against X-Ray](methodology.md) first for what F1,
+Read [How we score against X-Ray](methodology.md) first for what F1, precision,
-precision, recall, and misID mean in this report. All numbers below use the
+recall, and misID mean here. Every number below uses the per-second metric
-per-second metric
+([`scripts/optimizer/second_score.py`](https://REPOLINK/scripts/optimizer/second_score.py)):
-([`scripts/optimizer/second_score.py`](https://REPOLINK/scripts/optimizer/second_score.py)).
+the film is sampled once per second, and at each second the set of names the
+pipeline reports present is compared against Amazon X-Ray's scene cast for that
+second. X-Ray's ground truth is scene-level; the pipeline's output is per-second.
+That mismatch shapes every result.
 
-r50 (ArcFace w600k-R50) is excluded from the detailed comparison below. Its
+## The benchmark
-gallery was built with roughly 30% fewer reference images per actor than the
+
-other three models on the identical source photos (10808 vs 15055 total
+Nine films with public Amazon X-Ray scene data, all scored with the same
-embeddings across the same 2418 actors), which confounds any direct
+LVFace-B Glint360K gallery:
-comparison of its scores against the others. It remains in the
+
-[calibration curve comparison](best-model.md#first-signal-calibration-curves),
+Benny & Joon · Café Society · Downton Abbey: A New Era · Lord of War · Lovelace ·
-which does not depend on the training benchmark.
+The Many Saints of Newark · Scarface · Sound of Metal · Valerian.
+
+Two of these — Café Society and Scarface — are low-contrast, uniformly-graded
+films that break naive cut detection. They are deliberately kept in the benchmark
+because they are where the interesting failures live.
 
 ## Why replay makes this affordable
 
 Decoding video and running face detection, alignment, and embedding is the
-expensive part of this pipeline. Everything downstream of that (tracking,
+expensive part of the pipeline. Everything downstream — tracking, identity
-identity matching, scene aggregation) is cheap. KPN++'s node/network
+matching, scene aggregation — is cheap. KPN++'s node/network structure keeps those
-structure means those two stages are separate components connected by
+two halves as separate components joined by typed channels, so the expensive half
-typed channels, so the expensive stage can run once per film, cache its
+runs once per film and caches its output, and the cheap half can be re-run against
-output, and the cheap stage can be re-run against that cache as many times
+that cache as often as needed with different `Config` values.
-as needed with different Config values.
 
-`scene_analyze --dump-embeddings out.h5` runs the expensive half once per
+`scene_analyze --dump-embeddings out.h5` runs the expensive half once and writes
-film and writes per-frame face detections and embeddings to HDF5
+per-frame detections, embeddings, and (for the scene detector) per-frame RGB
-([`scripts/optimizer/SCHEMA.md`](https://REPOLINK/scripts/optimizer/SCHEMA.md)).
+histograms to HDF5. [`scripts/optimizer/replay.py`](https://REPOLINK/scripts/optimizer/replay.py)
-[`scripts/optimizer/replay.py`](https://REPOLINK/scripts/optimizer/replay.py)
+re-assembles the real C++ `face_tracker`, `identity_matcher`, and scene nodes into
-then re-assembles the real C++ `face_tracker`, `identity_matcher`, and
+a Python-driven KPN network and replays a film's cache through them, varying every
-`scene_tracker` nodes into a Python-driven KPN network and replays a
+tuning knob freely. No GPU inference and no video decode happen during a replay, so
-film's cached embeddings through them, varying `prob_threshold`,
+a full differential-evolution search over all nine films is tractable in one
-`anneal_sec`, `extinction_sec`, and `expand_gallery` freely. No GPU
+session rather than requiring re-encodes.
-inference and no video decode happen during a replay; each one completes
-in seconds. This is what makes a 512-evaluation differential-evolution
-search per model, per gallery mode, per expansion setting, tractable, and
-what made the full held-out validation across three models in this report
-possible in one session rather than requiring three full re-encodes of the
-benchmark set.
 
-`optimize.py` runs `differential_evolution` over this replay function as its
+Two concurrency limits are load-bearing and were paid for in wedged runs: replays
-objective, with DE-level parallelism (multiple candidate configs evaluated
+run at `DE_WORKERS=1` (concurrent DE candidates wedge the ROCm GPU), and each
-concurrently, each spawning its own replay subprocesses) on top of it. The
+candidate's per-film replays run at `REPLAY_WORKERS=8` with stderr discarded (the
-practical ceiling on this machine's GPU was 8 concurrent replay processes;
+replay sink's per-second prints otherwise flood the captured pipe and hang the
-9 silently degraded every score to 0.0% (well-formed output, wrong numbers,
+subprocess).
-not a crash), so `optimize.py` was run at `REPLAY_WORKERS=4 DE_WORKERS=2`.
 
-## Search space
+## The tuning knobs
 
-`popsize=10, maxiter=15` per combo (3 parameters, up to 512 evaluations,
+The opencv5 refactor replaced the old three-knob search with a **ten-knob**
-usually stopping earlier on DE's convergence tolerance).
+differential-evolution sweep. The knobs, and their shipped values:
-`anneal_sec`/`extinction_sec` bounds were widened from 1-30/1-15 to 1-60/1-60
-partway through the sweep. r50's 4 combos finished before the widening and
-used the old, narrower bounds; this is one more reason r50 is excluded from
-direct comparison here.
 
-## Training films and held-out films
+| knob | shipped | what it controls |
+| ---- | ------: | ---------------- |
+| `prob_threshold` | 0.485 | posterior P(match) above which a track is named |
+| `ownership_logodds` | 1.72 | log-odds a track needs before it produces presence |
+| `track_extinction_sec` | 31.0 | how long an idle track is held for re-detection |
+| `track_alpha` | 0.435 | tracker cost mix (0 = embedding only, 1 = spatial only) |
+| `evidence_rho_max` | 0.204 | evidence weighting ceiling |
+| `evidence_admit_below` | 0.784 | admit new evidence below this similarity |
+| `match_prior` | 0.433 | base-rate prior on a match |
+| `expand_band_lo` | 0.804 | low edge of the pose-expansion similarity band |
+| `expand_band_hi` | 0.952 | high edge of the pose-expansion band |
+| `presence_mode` | flood | track-extent vs scene flood-fill |
 
-9 films have dumped embeddings across all 4 models. 4 were used for
+The DE run over the first nine knobs (flood off, track-extent presence) converged
-optimization:
+at **64.0% macro F1** over 345 evaluations. Those values are the shipped
+[`src/config.hpp`](https://REPOLINK/src/config.hpp) defaults.
 
-- Café Society (62-cast)
+![10-knob presence sweep (Differential Evolution)](assets/images/de_search_landscape.png)
-- Lord of War (64-cast)
-- Scarface (67-cast)
-- Sound of Metal (14-cast)
 
-5 were held out, never seen by any optimizer run:
+The `track_extinction_sec` knob is worth calling out: at 31 s it holds an idle
+track alive for re-detection long enough to bridge an actor turning away or leaving
+frame briefly, without bridging across a genuine scene change. Getting this knob
+and the tracker/registry to agree on **one clock** (the evidence watermark, not
+wall-clock) was a correctness fix, not a tuning choice — before it, votes were
+silently dropped at the reap horizon.
 
-- Benny & Joon
+## The step change: flood-fill on learned boundaries
-- Downton Abbey: A New Era
-- Lovelace
-- The Many Saints of Newark
-- Valerian and the City of a Thousand Planets
 
-## Gallery coverage per film
+The 64.0% above is track-extent presence: an actor is reported only while an actual
+track is alive. **Flood-fill** instead reports an actor for the whole shot once
+they are seen in it — but that is only correct if the shot boundaries are good.
 
-The gallery has reference embeddings for 2418 actors, but coverage of any
+With the old grayscale cut detector as the boundary source, flood-fill barely beat
-given film's credited cast varies widely. This was previously reported as
+doing nothing (**64.0%**) and actively broke Scarface, where the detector fires
-one flat number (67% of X-Ray cast lacking a reference embedding, averaged
+once in 10,204 frames and flood then smears every actor across the whole film
-across the whole benchmark); the per-film breakdown is:
+(precision collapses to 26%).
 
-| film | cast credited | in gallery | coverage |
+The [learned scene-boundary detector](scene-boundary-detector.md) — an XGBoost
-|---|---|---|---|
+regressor over histogram-delta and audio features, with a per-film knee threshold —
-| Lord of War | 64 | 13 | 20.3% |
+fixes this. Macro per-second presence F1, at the shipped presence config:
-| Scarface | 67 | 15 | 22.4% |
-| The Many Saints of Newark | 48 | 13 | 27.1% |
-| Café Society | 62 | 17 | 27.4% |
-| Lovelace | 42 | 15 | 35.7% |
-| Valerian and the City of a Thousand Planets | 36 | 13 | 36.1% |
-| Benny & Joon | 23 | 12 | 52.2% |
-| Downton Abbey: A New Era | 36 | 22 | 61.1% |
-| Sound of Metal | 14 | 11 | 78.6% |
 
-Two training films (Lord of War, Scarface) have the worst coverage in the
+| boundary source for flood-fill | presence F1 |
-set, 20-22%. Their training-set F1 numbers below are partly capped by
+| ------------------------------ | ----------: |
-missing references, not purely by model quality. Downton Abbey has 61%
+| track-extent (flood off) | 62.6% |
-coverage, the second-best in the benchmark, yet the worst held-out recall
+| flood + grayscale cuts | 64.0% |
-of any film (39.4%, LVFace). Its recall problem is not primarily a coverage
+| **flood + learned detector (LOO)** | **74.9%** |
-problem; it is the extinction-bridging failure documented in the
-[LVFace deep dive](lvface-deep-dive.md#mechanism-1-extinction-bridging).
-Reproduce with `scripts/docs/gallery_coverage_per_film.py`.
 
-## Training results, 3 models × 2 gallery modes × 2 expansion settings
+![Macro presence F1 by flood-fill boundary source](assets/images/scene_presence_macro.png)
 
-Ranked by F1. misid = FPI_misid, the count of true wrong-actor
+The learned column is **leave-one-out**: each film is scored by a detector trained
-identifications (naming someone not in the film's cast at all), distinct
+on the other eight, so no film's presence is ever measured with a detector that saw
-from FPI, which also includes in-cast timing slips.
+it. That is the honest generalisation number, +12.3 points over track-extent, and
+**it improves every one of the nine films**.
 
-Each combo's row is its best **full-coverage** evaluation: the highest-F1 DE
+![Per-film presence F1 by boundary source](assets/images/scene_presence_by_source.png)
-evaluation in which all 4 training films replayed without a timeout (see
-[Dropped-film scoring](#a-scoring-bug-worth-recording-dropped-film-evaluations)
-below for why this qualifier is load-bearing and not the same as `argmax F1`
-over the raw sweep).
 
-| combo | F1 | P | R | TPI | FPI | misid | FN |
+| film | track-extent | flood+grayscale | flood+learned (LOO) |
-|---|---|---|---|---|---|---|---|
+| ---- | -----------: | --------------: | ------------------: |
-| LVFace-B_Glint360K_restricted_exp | 78.3% | 91.0% | 68.9% | 42830 | 3782 | 60 | 19492 |
+| Benny & Joon | 77.3 | 80.2 | 78.2 |
-| LVFace-B_Glint360K_restricted_noexp | 76.7% | 91.5% | 66.2% | 41149 | 3400 | 59 | 21173 |
+| Café Society | 59.1 | 62.2 | 69.8 |
-| arcface_w600k_mbf_restricted_exp | 76.2% | 90.0% | 66.2% | 64328 | 7480 | 0 | 33234 |
+| Downton Abbey | 41.0 | 51.8 | **78.6** |
-| arcface_r18_restricted_exp | 75.5% | 87.6% | 66.5% | 41399 | 5666 | 60 | 20923 |
+| Lord of War | 74.8 | 77.1 | 77.8 |
-| LVFace-B_Glint360K_full_exp | 75.3% | 89.7% | 65.4% | 47757 | 3407 | 232 | 26966 |
+| Lovelace | 70.3 | 74.0 | 78.2 |
-| arcface_w600k_mbf_restricted_noexp | 75.0% | 91.1% | 63.9% | 39752 | 3465 | 60 | 22570 |
+| The Many Saints of Newark | 37.5 | 43.9 | 53.4 |
-| arcface_r18_restricted_noexp | 73.5% | 91.3% | 61.7% | 38299 | 3220 | 60 | 24023 |
+| Scarface | 62.6 | **40.9** | **74.9** |
-| LVFace-B_Glint360K_full_noexp | 72.3% | 88.3% | 61.8% | 40363 | 3503 | 244 | 25850 |
+| Sound of Metal | 75.0 | 78.1 | 86.8 |
-| arcface_w600k_mbf_full_exp | 72.0% | 87.7% | 61.4% | 39875 | 3729 | 240 | 26338 |
+| Valerian | 65.6 | 67.7 | 76.2 |
-| arcface_w600k_mbf_full_noexp | 71.0% | 93.2% | 57.9% | 41699 | 2472 | 56 | 33024 |
-| arcface_r18_full_exp | 69.1% | 87.6% | 57.7% | 37342 | 3119 | 242 | 28871 |
-| arcface_r18_full_noexp | 66.6% | 91.3% | 53.1% | 34314 | 2362 | 107 | 31899 |
 
-![All combos ranked by training-set F1](assets/images/rep4_matrix_f1.png)
+The two headline films — Scarface (grayscale flood *breaks* it, learned flood on a
+film it never trained on takes it to 74.9%) and Downton Abbey (+37 points) — are
+the strongest evidence the detector generalises. See the
+[scene-boundary detector page](scene-boundary-detector.md) for the full story.
 
-The two clearest patterns: every model's best-scoring combo uses the
+We re-ran the ten-knob DE on top of the good boundaries to check whether the
-restricted gallery, and LVFace leads within both gallery modes. `full_exp`
+shipped config should change. It converged at 76.1% (+0.3 pp over the shipped
-(the shipped combination) is the best-scoring option that uses only
+config on learned boundaries) — inside the noise, not worth re-shipping. The
-features the running application currently supports; restriction is not
+boundaries, not the presence knobs, are where the win is.
-wired into the application yet (see
-[Whole vs. cast-restricted gallery](gallery-scope.md)).
 
-### A scoring bug worth recording: dropped-film evaluations
+## What the frames look like
 
-The numbers above are corrected ones. The raw `rep4_best_*.json` files, and an
+`scripts/optimizer/dump_error_frames.py` pulls representative seconds and draws
-earlier version of this table, reported a different `arcface_w600k_mbf_full_noexp`
+each face box against X-Ray's scene cast: **green** = true positive, **red** =
-row: **74.2% F1 at TPI 12645**, a third the TPI of every sibling combo. That was
+false positive (a name X-Ray does not credit to this scene — the real error),
-not a better config; it was an artifact of how the optimizer aggregates.
+**orange** = an unknown detection. Cast X-Ray lists as present but for whom no face
+was detected — the structural false-negatives a face pipeline can never box — are
+listed as a **blue** panel.
 
-`optimize.py` builds each candidate's score from only the films whose replay
+![A correctly identified second: green true-positive boxes](assets/images/lovelace_perfect_second.jpg)
-subprocess returned (`per_film = [m for m in ex.map(_one, films) if m is not
-None]`), then **averages** F1/precision/recall and **sums** TPI/FPI/misID over
-just those survivors. When a film's replay times out (the sweep ran near the
-8-process concurrency ceiling, so this happened intermittently), that film
-silently drops from both. A candidate whose hardest film timed out is therefore
-scored on an easier subset, and differential evolution, maximizing that score,
-will happily converge onto exactly such a candidate. For `mbf_full_noexp` the
-reported winner was one of 7 evaluations (out of 512) whose TPI had collapsed to
-a partial-film subset; its median-coverage evaluations sit around 51686 TPI.
 
-The fix here was to re-derive each combo's best row from its DE trajectory
+Every named frame in this documentation is regenerated against the current opencv5
-(`experiments/trajectories/rep4_*.jsonl`), keeping only evaluations within 30% of
+pipeline by [`scripts/scene_detector/rematch_frames.py`](https://REPOLINK/scripts/scene_detector/rematch_frames.py),
-that combo's median TPI (full 4-film coverage) before taking the best F1. This
+which auto-matches each example by film, actor, and class (TP/FP) so the images
-needs no re-running, the honest best configuration was already in the sweep,
+never drift from the shipped behaviour. Where the current pipeline no longer makes
-just not the one `argmax F1` selected. Three combos moved: `mbf_full_noexp`
+a July-era error — the Zooey Deschanel misID in Many Saints is the clearest case —
-74.2% → **71.0%**, `LVFace_full_noexp` 72.4% → **72.3%** (and its misID, 0 → 244,
+the frame is dropped rather than staged, because the improvement is real.
-was itself a dropped-film artifact), `mbf_restricted_exp` 76.5% → **76.2%**. The
-shipped LVFace `full_exp` winner was unaffected, its reported evaluation already
-had full coverage (TPI 47757 ≈ median). `experiment_charts.py` applies the same
-`clean_best` filter, so every figure on this page matches the corrected table.
-The underlying `optimize.py` aggregation is also being fixed so a dropped-film
-evaluation can never be selected as a winner again.
 
-### Per-film training breakdown
+## The structural recall ceiling
 
-The 75.3% LVFace training figure is a macro average across 4 films, not a
+Precision against X-Ray is near-perfect on identified faces; recall is capped by
-uniform result:
+two things the pipeline cannot fix:
 
-| film | LVFace F1 | mbf F1 | r18 F1 | best model |
+1. **X-Ray credits people whose faces never appear on screen** in a scene — voice,
-|---|---|---|---|---|
+   back-of-head, or simply off-camera cast. No face pipeline can box a face that is
-| Café Society | 68.1% | 62.2% | 60.1% | LVFace |
+   not there. These are the blue-panel names.
-| Lord of War | 75.6% | 77.2% | 75.6% | mbf |
+2. **Gallery coverage.** A large fraction of X-Ray cast has no reference image in
-| Scarface | 71.5% | 68.6% | 64.1% | LVFace |
+   the gallery, so those actors can never be matched regardless of detection. This
-| Sound of Metal | 78.8% | 76.5% | 71.6% | LVFace |
+   is the dominant remaining recall limiter and is addressable by fetching more
+   reference photos, not by tuning.
 
-LVFace does not win every training film. mbf scores higher on Lord of War
+Both are documented in [how we score against X-Ray](methodology.md).
-(77.2% vs 75.6%). LVFace's own training-film range is 68.1% to 78.8%, a
-10.7pp spread, smaller than the 37pp spread seen on held-out films but real.
-Reproduce with `scripts/docs/run_holdout_all_models.py --films training`.
 
-## Held-out validation, all 3 models
+## In the pipeline
 
-The training matrix above is training-set fit. Each model's own tuned
+The learned detector runs live inside `scene_analyze` as a post-EOF step (the
-`full_exp` config was replayed against the 5 held-out films, scored the
+per-film knee needs every peak, so it can only run once the whole film is seen).
-same way:
+XGBoost inference is built into the binary via CMake (`SAE_SCENE_XGB`); the audio
-
+log-PSD uses FFTW on the existing FFmpeg decode. The shipped model is trained on
-| film | LVFace F1 | mbf F1 | r18 F1 |
+the **C++-extracted** features so training and inference share one implementation.
-|---|---|---|---|
+Verified end to end through `scene_analyze` on a movie file and through the Jellyfin
-| Benny & Joon | 83.0% | 78.5% | 77.1% |
+work-queue worker.
-| Lovelace | 77.5% | 73.7% | 72.2% |
-| Valerian and the City of a Thousand Planets | 74.1% | 70.2% | 71.0% |
-| Downton Abbey: A New Era | 56.2% | 55.0% | 53.0% |
-| The Many Saints of Newark | 46.3% | 44.5% | 42.1% |
-| **macro average** | **67.4%** | **64.4%** | **63.1%** |
-
-LVFace scores highest on every one of the 5 held-out films; the ranking
-never flips. Total misIDs across the 5 films: LVFace 1032, mbf 2197, r18
-1224. LVFace has less than half mbf's misID count while also scoring
-higher on every film. This directly confirms the model choice out of
-sample; it is not inferred from the training numbers alone. See the
-[LVFace deep dive](lvface-deep-dive.md) for frame-level detail on where and
-why LVFace still fails on the two worst films. Reproduce with
-`scripts/docs/run_holdout_all_models.py`.
-
-## Two effects in isolation: gallery scope and pose expansion
-
-Averaging across the 3 compared models (r50 excluded) isolates each variable
-from model choice.
-
-**Gallery scope**, averaged over both expansion settings and all 3 models
-(6 evaluations per row):
-
-| scope | F1 | P | R | total misID |
-|---|---|---|---|---|
-| full | 71.1% | 89.6% | 59.6% | 1121 |
-| restricted | 75.9% | 90.4% | 65.6% | 299 |
-
-Restriction improves every metric at once. This is not a precision/recall
-trade: +4.8pp F1, +6.0pp recall, and roughly a quarter the misIDs. Fewer
-candidates in the matcher's search space means fewer opportunities for a
-lookalike false match, and the recall gain shows this does not cost real
-detections. Restriction is currently an offline optimizer technique, not a
-runtime feature of the application; see
-[Whole vs. cast-restricted gallery](gallery-scope.md) for what building it
-into the application would require.
-
-**Pose expansion** (promoting a confidently identified track's novel-pose
-views into a per-film gallery annex,
-[`src/gallery/track_gallery.hpp`](https://REPOLINK/src/gallery/track_gallery.hpp)):
-
-| scope | expansion | F1 | R | misID |
-|---|---|---|---|---|
-| full | off | 70.0% | 57.6% | 407 |
-| full | on | 72.1% | 61.5% | 714 |
-| restricted | off | 75.1% | 63.9% | 179 |
-| restricted | on | 76.7% | 67.2% | 120 |
-
-In restricted mode, expansion is a clean win: +1.6pp F1, +3.3pp recall,
-misID drops. The annex only competes against the film's own roughly 15-actor
-cast, so a new pose of a known actor is unlikely to be confused with someone
-else. In full mode, expansion buys +2.1pp F1 and +3.9pp recall but at a real
-cost: misID rises from 407 to 714 as the same new-pose view now competes
-against the full 2418-actor gallery, where a confidently learned pose is more
-likely to match the wrong person. On the full gallery it is a recall-vs-misID
-trade, not a free gain. This training-set effect
-did not reproduce on held-out data; see
-[Does pose expansion help?](pose-expansion.md) for the full held-out test
-and the two methodology bugs caught while checking it.
-
-## Calibration curves
-
-Each gallery carries a fitted Platt sigmoid `P(match | sim) = σ(a·sim + b)`,
-stored directly in the gallery HDF5
-([`src/gallery/gallery_calibration.hpp`](https://REPOLINK/src/gallery/gallery_calibration.hpp)).
-This measures discriminative power independent of whatever
-`prob_threshold` a given run used:
-
-![Calibrated P(match|similarity) for all four models](assets/images/calibration_curves.png)
-
-LVFace has the steepest curve (`a=17.7` vs 15.3-16.2 for the ArcFace
-variants) and the lowest P=0.5 decision boundary (similarity 0.23 vs
-0.27-0.31), separating same-actor from different-actor pairs more
-confidently at a lower similarity than any ArcFace variant tested,
-including r50. Generated by
-[`scripts/docs/calibration_chart.py`](https://REPOLINK/scripts/docs/calibration_chart.py).
-
-## Extinction and anneal window search
-
-Every one of the 512 DE evaluations for the winning LVFace `full_exp`
-combo, plotted over the `prob_threshold` × `extinction_sec` plane:
-
-![DE search landscape: 512 evaluations over prob_threshold × extinction_sec](assets/images/de_search_landscape.png)
-
-Nearly everything scoring well sits at `extinction_sec` above 50, across a
-wide range of thresholds. Short extinction windows are uniformly weaker:
-under a strict threshold, there is no good configuration in that region of
-the search space. The optimizer converged with `anneal_sec=59.2,
-extinction_sec=59.2`, about 99% of the widened 60s bound, which raises an
-open question not resolved in this round: does performance keep improving
-past 60s, or does it plateau there. Not chased further this pass.
-
-## Caveats
-
-- r50's 4 combos used the older, narrower search bounds (1-30/1-15 instead
-  of 1-60/1-60) and are further confounded by its thinner gallery. Excluded
-  from all comparisons above except calibration.
-- The shipped defaults use `full_exp` (75.3% training F1), not the
-  higher-scoring `restricted_exp` (78.3%), because cast restriction is not
-  a runtime feature of the application yet.
-- `expand_gallery` is mode-dependent, not a free win. Averaged across models
-  on the full gallery it trades misIDs for recall (see the pose-expansion
-  table). For LVFace specifically, though, `full_exp` beats `full_noexp` on
-  every axis at once (F1 75.3 vs 72.3, precision 89.7 vs 88.3, recall 65.4 vs
-  61.8, misID 232 vs 244), so the shipped `full_exp` is a clean choice for
-  this model, not an F1-vs-safety trade. (An earlier version of this page
-  reported `full_noexp` at 72.4% with zero misIDs and higher precision, which
-  made it look like the safer option; that was the dropped-film artifact
-  described above, not a real property of the config.)
-- Switching the default model is an operational change: any gallery built
-  from a different model's embeddings must be rebuilt before the new
-  default takes effect.
-
-## Reproduce
 
 ```bash
-# 4-film training matrix, all 4 models × 2 gallery modes × 2 expansion settings
+scene_analyze --movie <file> --gallery <gallery.h5> \
-bash experiments/run_rep4_subprocess.sh
+    --scene-xgb-model models/scene_boundary_xgb.json
-
-# single combo
-SAE_EXPAND=1 REPLAY_WORKERS=4 DE_WORKERS=2 python3 scripts/optimizer/optimize.py \
-    --manifest experiments/manifests/rep4_LVFace-B_Glint360K_full.json \
-    --gallery experiments/galleries/gallery_LVFace-B_Glint360K.h5 \
-    --params prob_threshold:0.5:0.999 anneal_sec:1:60 extinction_sec:1:60 \
-    --popsize 10 --maxiter 15 --trajectory traj.jsonl --out best.json
-
-# held-out validation, all 3 models, 5 films
-python3 scripts/docs/run_holdout_all_models.py --out docs_data/holdout_all_models.json
-
-# per-film training breakdown, all 3 models, 4 films
-python3 scripts/docs/run_holdout_all_models.py --films training --out docs_data/training_per_film.json
-
-# gallery coverage per film
-python3 scripts/docs/gallery_coverage_per_film.py --out docs_data/gallery_coverage_per_film.json
-
-# regenerate this page's charts from experiments/ artifacts
-python3 scripts/docs/experiment_charts.py --out-dir docs/assets/images
-
-# one frame per distinct out-of-cast name across all 9 films (used in the deep dive)
-python3 scripts/docs/first_fpi_frames.py
 ```
 
-See also the session log
+## Reproducing the benchmarks
-[`experiments/SESSION_STATE.md`](https://REPOLINK/experiments/SESSION_STATE.md).
+
+Gallery `.h5` files, embedding dumps, the X-Ray corpus, and DE trajectories are not
+committed. They are pushed to the Gitea package registry and pulled on demand:
+
+```bash
+scripts/artifacts/pull_artifacts.sh galleries
+scripts/artifacts/pull_artifacts.sh experiment-data
+
+# per-second audio features, C++ feature matrices, train + downstream A/B
+.venv-rocm/bin/python scripts/scene_detector/extract_audio_features.py \
+    --manifest experiments/manifests/films_LVFace_opencv5.json
+.venv-rocm/bin/python scripts/scene_detector/train_xgb_cpp.py --train-all
+scripts/scene_detector/downstream_presence.py
+```

docs/pose-expansion-2026-07.md

@@ -1,3 +1,5 @@
+> **Archived (July 2026).** This report covers the pre-opencv5 framework and the 4-model ArcFace/LVFace bake-off. It is superseded by the current [experiment log](model-bakeoff.md) for the opencv5 build. Kept for provenance; the numbers here are historical.
+
 # Pose expansion: does promoting new poses mid-film help?
 
 `expand_gallery`
@@ -12,7 +14,7 @@ in the same film, without touching the baked gallery.
 
 Averaged across the 3 compared models (r50 excluded), on the 4 films used
 for optimization. These are the corrected, full-coverage figures, see the
-[dropped-film note](model-bakeoff.md#a-scoring-bug-worth-recording-dropped-film-evaluations)
+[dropped-film note](model-bakeoff-2026-07.md#a-scoring-bug-worth-recording-dropped-film-evaluations)
 in the experiment log for why an earlier version of this table overstated the
 full-mode misID jump (209 → 864) that was itself partly a truncation artifact:
 
@@ -26,7 +28,7 @@ full-mode misID jump (209 → 864) that was itself partly a truncation artifact:
 In restricted mode, expansion looks like a clean win: +1.6pp F1, +3.3pp
 recall, lower misID. In full mode it looks like a recall-for-misID trade:
 +2.1pp F1, +3.9pp recall, but misID rises from 407 to 714. See
-[the full experiment log](model-bakeoff.md) for the per-model breakdown.
+[the full experiment log](model-bakeoff-2026-07.md) for the per-model breakdown.
 This asymmetry motivated the question below: does turning expansion on
 change what gets recognized frame by frame, or is the aggregate F1 shift
 coming from something else.
@@ -105,6 +107,6 @@ contribution, such as tagging which reference embedding won each match;
 neither was in scope for this pass.
 
 Do not treat the training-set exp/noexp numbers in
-[the full experiment log](model-bakeoff.md) as proof that expansion changes
+[the full experiment log](model-bakeoff-2026-07.md) as proof that expansion changes
 real-world behavior in either direction. On the evidence gathered so far,
 it does not move the needle enough to see.

docs/requirements.md

@@ -31,7 +31,7 @@ Status: `Done` · `In Progress` · `Planned` · `TBD` · `Withdrawn`
 | AR-001 | Detect faces in sampled frames; emit bbox, confidence, 5-point landmarks in original pixel space | SR-002 | High | Done |
 | AR-002 | Minimum face size **40×40 px** (VR-013 measured end to end; VR-005's 32 px is an embedder-only upper bound), expressed in **original** resolution (decoupled from `dense_scale`) | SR-002 | High | **Done** — `FaceDetectorFunc::drop_undersized()`. The threshold is divided by `bbox_upscale` rather than every box multiplied, which keeps the comparison on the detector's own numbers and means turning `dense_scale` on cannot silently raise the minimum face the pipeline accepts. Verified at the threshold and at `dense_scale` 0.5 (UT-002), and end to end on the fixture (IT-001) — the superhero dump's smallest side is *exactly* its recorded 32 px, so the filter is binding there rather than vacuously satisfied |
 | AR-003 | No fixed per-frame face cap — crowd scenes must not lose background cast | SR-002 | Medium | **Done** — `max_faces` defaults to 0 (no cap); the matcher batches through its GEMM buffer instead of throwing |
-| AR-004 | Backpressure: unbounded faces/frame absorbed by slowing, never by dropping or throwing | SR-002 | High | **Mostly** — node outputs *park* on a full channel: the value is held, the worker released, and a channel space-callback resumes the node. Replaces `push_blocking`, which parked a scheduler worker inside the push and, with one thread per node, stopped that node draining its own input. Verified: 385/385 frames, 0 drops. Two remaining holes now closed: **(a)** `FanoutNode` dropped on overflow rather than waiting, so the AR-010 scene join shed frames exactly when the dense branch fell behind — measured at **9 of 2192 items delivered** to the slower of two branches, now lossless with the fast branch throttled to within its buffering; **(b)** the residual hang, recorded as ~1 run in 20 at a 300 s timeout, was a **startup** lost wake, not a mid-stream one — `start()` enables a node's inputs several statements before it installs the push callback, and a producer firing into that gap is accepted by the ring while waking nobody, since `Channel::push` signals only the empty→non-empty edge. Signature is zero items delivered, never a partial stall. Reproduced 7 times in 24 under CPU contention and 0 in 10 without; `start()` now closes with the level-triggered `on_input_ready()`, giving 0 in 24 on the same harness. **Consequence to hold onto:** a lossless fanout makes join depth a correctness precondition — one branch can now run ahead of another only by the slower branch's buffering, so `kSceneJoinDepth` must exceed the TransNetV2 window. **Gap:** capacity is still counted in *items*, not bytes, so a crowd frame carrying 60 crops occupies one slot exactly as an empty one does — the memory ceiling the plan asks for is unenforced |
+| AR-004 | Backpressure: unbounded faces/frame absorbed by slowing, never by dropping or throwing | SR-002 | High | **Mostly** — node outputs *park* on a full channel: the value is held, the worker released, and a channel space-callback resumes the node. Replaces `push_blocking`, which parked a scheduler worker inside the push and, with one thread per node, stopped that node draining its own input. Verified: 385/385 frames, 0 drops. Holes closed since, in the order they surfaced: **(a)** `FanoutNode` dropped on overflow rather than waiting, so the AR-010 scene join shed frames exactly when the dense branch fell behind — measured at **9 of 2192 items delivered** to the slower of two branches, now lossless with the fast branch throttled to within its buffering; **(b)** the residual hang, recorded as ~1 run in 20 at a 300 s timeout, was a **startup** lost wake, not a mid-stream one — `start()` enables a node's inputs several statements before it installs the push callback, and a producer firing into that gap is accepted by the ring while waking nobody, since `Channel::push` signals only the empty→non-empty edge. Signature is zero items delivered, never a partial stall. Reproduced 7 times in 24 under CPU contention and 0 in 10 without; `start()` now closes with the level-triggered `on_input_ready()`, giving 0 in 24 on the same harness — though the *cause* was narrower than recorded there and is fixed properly in **(e)**; **(c)** `FilterNode` and `RouterNode` were the last data paths still using the throwing `push()` with the exception swallowed, so a full output discarded the value — including the **EOF sentinel**. The decimator passes EOF by predicate (`if (f.eof) return true;`) but its output is reliably full, the embedder being the slowest node in the chain, so the token was discarded, nothing downstream ever shut down, and the run had to be killed. **This is the wedge.** Both now route sentinels out-of-band and retry data until taken; the regression case delivers 6 of 40 values and never sets `saw_eof` before, 40 and terminating after; **(d)** the sentinel could be delivered *ahead of* a value still queued behind it — `pop()` observed the ring empty and then took the sentinel, and a producer can push a value *and* publish the sentinel inside that window, so any consumer treating EOF as a hard stop loses the tail. `take_sentinel` now re-checks emptiness *after* observing `has_eof_`, which is sound because the sentinel is published with a release store after the ring pushes. ~1 run in 15 before, 0 in 25 after; **(e)** two `fire_once` invocations for one node could overlap, because the submit gate was released before the firing had finished touching node state. That breaks the one-slot park the whole scheme rests on — a parked value can be overwritten by the other firing, with no drop recorded anywhere. ThreadSanitizer caught it as a race on `pending_done_`; the release is now the last act of a firing. The same sweep found the callbacks themselves being written while a running neighbour read them (ten TSan races), which is the *actual* cause of the startup lost wake in **(b)** — callbacks are now installed in a `prepare()` pass before any node starts. **New constraint:** a channel carries at most **one undelivered sentinel**; a second offered before the first is taken is refused and reported, never queued and never overwritten, since two control tokens on one channel means the stream ended twice. Single-shot EOF today, live the moment a pipeline is reused for a second input. **Consequence to hold onto:** a lossless fanout makes join depth a correctness precondition — one branch can now run ahead of another only by the slower branch's buffering, so `kSceneJoinDepth` must exceed the TransNetV2 window. Making the decimator lossless also makes it a backpressure point rather than a relief valve: the source now throttles to the face branch instead of quietly thinning it. Correct under this requirement, but it changes the shape of a loaded run and is **not yet benchmarked**. **Gap:** capacity is still counted in *items*, not bytes, so a crowd frame carrying 60 crops occupies one slot exactly as an empty one does — the memory ceiling the plan asks for is unenforced |
 | AR-005 | Align to 112×112 via ArcFace 5-point similarity transform, fitted by **Umeyama least squares over all five points** (as InsightFace does) — never a robust fit, which would discard the landmarks AR-030 reads | SR-002 | High | **Done** — `umeyama_similarity()`. The RANSAC fit it replaces disagreed by a median 17 source px on 400 headshots, 83.5% of crops embedding below cos 0.99, and was unstable and RNG-driven: rebuilding caught 1614 near-duplicates against the original build's ~100. **All galleries rebuilt** (2456 actors, 10254 embeddings); measured separation gain is small (0.583 → 0.590), so recorded accuracy figures should be re-run but are not expected to move far |
 | AR-006 | 512-d L2-normalised embeddings, batched | SR-002 | High | Done |
 | AR-007 | Associate detections by IoU + embedding, with **frame-dependent** weighting | SR-002 | High | **Done** — `track_alpha` is the base for ordinary frames; drops to embedding-only on cut/boundary and for dormant tracks |
@@ -104,7 +104,7 @@ Status: `Done` · `In Progress` · `Planned` · `TBD` · `Withdrawn`
 | ID | Requirement | Traces to | Priority | Status |
 |---|---|---|---|---|
 | VR-001 | HDF5 post-inference dump at the embedded-frame boundary | PR-002 | High | Done |
-| VR-002 | Replay drives the **real** KPN nodes, not a reimplementation | PR-002 | High | **In Progress, and worse than it looked.** The C++ side is real (`tests/test_replay_fixtures.cpp`, determinism asserted). The *Python* side is not runnable: `sae_kpn` has not compiled since the AR-007/AR-008 redesign — the binding builds `FaceTrackerFunc` from a `Config` alone, and the tracker has required a registry and a calibration since. Any `.so` in a stale `build/` predates that. Now behind `SAE_BUILD_KPN_BINDINGS=OFF` so the breakage is attributed rather than rediscovered; fixing it is VR-011. **Also correct the fixture claim:** the dumps are *not* committed (`tests/fixtures/dumps/.gitignore`) — they are Gitea package-registry artifacts, pulled by the CI job |
+| VR-002 | Replay drives the **real** KPN nodes, not a reimplementation | PR-002 | High | **Done** — including the sink, as of VR-011. Worth recording what the reimplementation was hiding: `build_minimal` rebuilt windows in Python from per-frame annotations, which never consult the registry, so it kept producing plausible output while registry-based presence in replay was returning **nothing at all**. The first run of the real chain emitted 0 actors on a film where 1647 frames carried an identified face. A reimplementation does not merely risk disagreeing with the pipeline; it can conceal the pipeline being broken |
 | VR-003 | Scoring: micro-F1 against X-Ray, precision/recall logged at every evaluation | PR-002 | High | Done |
 | VR-004 | Reproducible validation corpus with ground truth | PR-002 | High | Done |
 | VR-005 | Minimum face size study — TPI/FPI vs probe size, gallery held at native res | PR-002 | Medium | **Done** — knee at 24–32 px; 32 px gives 98.1% TPI, 0.0 FPI at every size. Degrades an already-aligned 112×112 crop, so it isolates the embedder and is an **upper bound**; VR-013 measures the same question end to end and AR-002 takes its number, not this one |
@@ -113,12 +113,13 @@ Status: `Done` · `In Progress` · `Planned` · `TBD` · `Withdrawn`
 | VR-008 | Gallery scaling benchmark — throughput vs gallery size | PR-002 | Medium | Planned |
 | VR-009 | Verify accumulated posteriors are calibrated against held-out tracks | PR-002 | High | Planned |
 | VR-010 | Dump provenance attributes — embedder model, detector settings, `dense_scale`, `scene_detect`, sample rate | PR-002 | **High** | **Done** — `DumpProvenance` in `embedding_dump_node.hpp`, written as root attributes and read back tolerantly. Every field is optional so a pre-VR-010 dump reads as *unknown* rather than as a default; a silently-defaulted `detector_conf` is exactly the fabricated provenance this exists to prevent. This row said `Planned` while five VR-010 tags sat in the code — stale in the opposite direction to the rest of this register |
-| VR-011 | Rewrite the replay harness for the post-AR-012 output contract | PR-002 | High | Planned |
+| VR-011 | Rewrite the replay harness for the post-AR-012 output contract | PR-002 | High | **Done** — `sae_kpn` compiles again and the replay drives the whole chain including `ResultSinkFunc`, so presence comes from `TrackRegistry` claims rather than being rebuilt in Python. The three per-node factories are replaced by one `add_pipeline` that mirrors `main.cpp`'s construction order — the ordering constraint (matcher fits the calibration, registry needs a discounter from it, tracker needs both, sink needs the claims) is what a factory-per-node API could not express, and is why the tracker factory kept building `FaceTrackerFunc{cfg}` against a signature that had stopped existing. `build_minimal` and `anneal_sec` are gone. Verified end to end on the SuperHero fixture: 5 actors, 32 windows, 0 dropped votes |
 | VR-012 | Quality-knee study — TPI/FPI vs sharpness and vs pose, as VR-005 did for size; also settles whether the 5-point pose proxy needs a dedicated landmark model | PR-002 | Medium | Planned |
 | VR-014 | Audio-signature **offset recovery on real content** — a known trim recovered from film audio, not from the synthetic golden tone | PR-002 | Medium | **Done** — 40 random in-cap offsets, every one recovered to the nearest frame: **worst error 46 ms against a 500 ms budget**, and 46 ms is the floor rather than a result, since the offset is quantised to whole 92.88 ms frames. The `runtime/2` anchor confirmed through real head-trimmed files (a `delta` trim moves the window by `delta/2`). The one soft spot is **tier labelling, not accuracy**: the score falls with sub-frame misalignment (0.94–0.99 near a frame boundary, 0.69–0.73 at half a frame), so 27/40 correct alignments were demoted to `loose`. ±1 frame of slack in the *score* fixes it — measured, all 40 back to `audio` (min 0.906), false matches unmoved at 0.12–0.16, costing 81 ms of the budget |
 | VR-015 | Per-node cost and bottleneck attribution for a run — where the time actually goes | PR-004 | High | **Done** — `--benchmark <path>` on `scene_analyze`; `src/benchmark.hpp`. Reports cumulative CPU and wall time per node, and locates the pacing node from sampled channel occupancy rather than from time-in-node, which backpressure inflates. Verified UT-120…UT-124 |
 | VR-013 | Cross-source identification probe — gallery from one recording, probes from another, swept over input resolution end to end | PR-002 | Medium | **In Progress** — holding 90% of the plateau needs ~50 px end to end against VR-005's ~22 px, the gap being detection and landmark error; **`min_face_px` 40, since 32 admits faces in the falling region** (AR-002). FPI 0.0% at every scale. Ceiling is cross-view, not resolution |
 | VR-016 | **Cut-detection cadence study** — is `cut_threshold` 0.70 meaningful at the rate `camera_pos` is actually fed? | PR-002 | Medium | **Planned.** The histogram cut detector is the one always-on signal with no recorded provenance, and its input rate is not the rate it was fitted at. With `--scene-detect` off, `camera_pos` sits downstream of a source already decimated to `sample_fps`, so at the 1.0 default it compares frames **one second apart** — inside a single shot those differ enormously, and 0.70 correlation is a low bar to clear. With `--scene-detect` on it sees native-rate frames instead, so the same constant means two different things depending on an unrelated flag. This is AR-011's argument ("every model gets the input it was trained for") applied to a non-neural detector, and it matters because `is_cut` drives `track_alpha` to 0 and clears every expansion buffer. Cheap first measurement: run `camera_pos` over a `hero/` clip at 1/2/5 fps and compare cut counts against `tests/fixtures/dumps/scene_bounds.json`. The committed 5 fps dump shows 2.6% of frames flagged; nobody has measured 1 fps |
+| VR-017 | **Vote-lag study** — how often does the matcher fall more than `track_extinction_sec` behind the tracker on real content? | PR-002 | **High** | **Planned.** Channel depth is a correctness parameter between `face_tracker` and `identity_matcher`, and the constraint runs opposite to the scene join's: there `kSceneJoinDepth` must EXCEED the TransNetV2 window, here the depth must be UNDER `track_extinction_sec × sample_fps`. Backpressure is what makes it bite — it is working, and a lossless channel converts depth into lag by design. Both nodes are 16 deep in `main.cpp`, which at the default `sample_fps` 1.0 is ~16 s of lag against a 5 s window, so `scene_analyze` can drop identity votes and until now said nothing. It now reports `dropped_votes` at shutdown; this row is the measurement that decides whether that should be fatal, and whether the right fix is bounding the depth or removing the coupling (reap on the matcher's clock rather than the tracker's, so a vote cannot be late by construction) |
 
 ---
 
@@ -314,7 +315,7 @@ because it will be trusted.
 | AR-001 | T3 | Detector returns plausible boxes on a known frame | — smoke only |
 | AR-002 | T2 | Faces below 40 px (original res) are dropped | Exactly at threshold; with `dense_scale` 0.5 — the interaction that motivated the requirement |
 | AR-003 | T2 | No cap applied; a 40-face frame yields 40 | Crowd frame |
-| AR-004 | T1 | Saturated input blocks rather than drops or throws | Bounded queue at capacity; **byte-based** limit with large crops; SIGTERM mid-block. Two cases the KPN suite now pins, both of which failed before being written: a fanout feeding an unequal pair loses nothing *and* throttles the fast branch (either assertion alone passes on a broken implementation); and a node started with data already in its input still fires — the startup lost wake, which needs no contention to reproduce once the state is constructed directly |
+| AR-004 | T1 | Saturated input blocks rather than drops or throws | Bounded queue at capacity; **byte-based** limit with large crops; SIGTERM mid-block. Cases the KPN suite now pins, each of which failed before being written: a fanout feeding an unequal pair loses nothing *and* throttles the fast branch (either assertion alone passes on a broken implementation); a filter delivers EOF into a saturated output; a sentinel is never delivered ahead of a queued value; a twice-parked value keeps its payload; and a node started with data already in its input still fires — the startup lost wake, which needs no contention to reproduce once the state is constructed directly |
 | AR-005 | T1 | Known landmarks → expected 112×112 warp; the fit never mirrors | Landmarks near frame edge; degenerate/collinear points; a mirrored set — SVD returns a reflection unless the determinant guard rejects it |
 | AR-006 | T3 | Embeddings are unit-norm | Batch smaller than, equal to, larger than `embed_batch_size` |
 | AR-007 | T2 | Association picks the right track | Two faces crossing paths; one leaving frame as another enters |

docs/scene-boundary-detector.md

@@ -0,0 +1,205 @@
+# The learned scene-boundary detector
+
+Presence uses **flood-fill**: an actor seen once inside a shot is reported for the
+whole shot (`[prev_boundary, next_boundary]`). That only works if the boundaries
+are good. This page is the story of getting them good — a learned scene-boundary
+detector that lifts per-second actor-presence F1 from **62.6% to 74.9%** across
+the nine-film X-Ray benchmark, and fixes the film where naive flood-fill was
+actively harmful.
+
+That 74.9% is the **leave-one-out** figure: each film is scored by a detector
+trained on the *other eight*, so no film's presence is measured with a detector
+that ever saw it. It is the honest generalisation number, and it is only ~1 point
+below the all-nine-trained model (75.8%) — the detector barely overfits.
+
+## Why the old cut detector wasn't enough
+
+The always-on boundary source was the grayscale histogram-correlation cut detector
+(`camera_position_change_detector`): mark a cut when the frame-to-frame grayscale
+histogram correlation drops below 0.70. It is cheap and it fires on obvious hard
+cuts, but on a low-contrast, uniformly-graded film it is nearly blind. On
+**Scarface** it fired **once in 10,204 frames**. Flood-fill then snapped every
+actor across essentially the whole film:
+
+| Scarface | precision | recall |
+| -------- | --------- | ------ |
+| flood + grayscale cuts | **26%** | 95% |
+| track-extent (no flood) | 92% | 45% |
+
+That single failure is what motivated everything below: flood-fill needs a
+boundary source that works regardless of grade.
+
+## What we are detecting, and why it is hard
+
+The training target is **Amazon X-Ray scene boundaries** (`scenes.csv`). These are
+*narrative* scenes — a new location or beat in the story — not shot cuts. There
+are only ~20–60 of them per film (median scene ~170 s), and many transition
+*within* continuous visual style and continuous audio. So the signal is sparse and
+often genuinely faint: a boundary detector working from audio-visual features can
+never recall a narrative cut that has no audio-visual signature.
+
+This shapes every result: absolute boundary-F1 is modest by construction. What
+matters is the **downstream** number — does snapping flood-fill to these
+boundaries name the right actors — and there the gain is large.
+
+## The features (what worked, measured)
+
+Everything is per second, aligned to the 1-fps presence grid.
+
+- **Delta histograms, not raw histograms.** The raw RGB histogram encodes what a
+  frame *looks like*, not that it *changed* — measured boundary separability ~1.4×.
+  The **symmetric histogram delta** `|hist(t+k) − hist(t−k)|` separates boundaries
+  **4–5×**. Leading with deltas (k = 1,2,4,8 s) and dropping the raw histogram was
+  the single biggest feature win (LSTM F1 7.5% → 10.8%).
+- **A multi-scale "ramp" bank.** Antisymmetric matched filters at half-widths
+  H = 2,4,6,8,10 s; the model weights the scales. Different films' boundaries peak
+  at different widths.
+- **A time-since-last-boundary "debounce" clock**, scaled by the corpus mean scene
+  length (~205 s), encoding that scenes don't restart moments apart.
+- **Audio log-PSD** (per-second, 4 s window, ~57 log-frequency bins). Measured
+  weak on its own — a standalone audio cutter scored only 3–6% held-out F1, because
+  narrative boundaries usually have continuous audio — but it is complementary on
+  the films where video is weak (Downton, Sound of Metal), so it is included and
+  the model uses it where it helps.
+
+![Detector development at strict ±2 s tolerance, and where the shipped detector landed at the ±20 s tolerance the pipeline uses](assets/images/scene_detector_evolution.png)
+
+The left panel is the *feature* development, scored at a strict ±2 s tolerance so
+each change is visible — this is where "delta beats raw histogram" was measured, not
+the shipped tolerance. The right panel is the shipped detector at the ±20 s
+tolerance the pipeline actually uses (see below). The two panels are on different
+tolerances by design and must not be read as one curve.
+
+Dead ends, all measured and discarded: audio-only detection; raw
+histograms/PSDs as input; a two-tower BiLSTM (no better than the tree, far slower);
+larger FFT windows / more frequency bins (worse — boundaries are short events);
+and TransNetV2 (a Conv3D net that will not co-reside with the ROCm/VAAPI stack).
+
+## The model
+
+- **XGBoost regressor** over a ±3 s window of the features above, predicting a
+  **soft Gaussian proximity-to-boundary target** (`exp(-(d/σ)²)`, σ = 10 s).
+  Regression to a soft target — rather than a hard 0/1 label — stops a near-miss
+  from being trained as a hard negative, and yields a smooth score whose **peaks**
+  are the boundaries.
+- **Per-film knee threshold.** The predicted peak heights form a
+  convex-decreasing curve; the knee (max drop below the endpoints' chord) is where
+  real boundaries give way to noise. Selecting at the knee **self-calibrates the
+  boundary count** to roughly the true scene count, per film, with no global
+  threshold that would be wrong for every grade.
+- **Trained on all nine films** for the shipped model. Keeping the low-contrast
+  grades (Café Society, Scarface) in training matters most: on its own training
+  films the shipped model reaches **72.9% macro boundary-F1** (per-film 51–86%),
+  versus **29.8%** for the grayscale baseline on the same films.
+
+Boundary detection, held out (leave-one-out, ±20 s tolerance — appropriate given
+~170 s scenes): **44.1% macro F1, versus 29.8% for the grayscale baseline** — the
+honest generalisation number, each film scored by a detector trained on the other
+eight. Even the low-contrast grades generalise (Scarface held out 32%, Café Society
+51%), where the grayscale detector scores 0% and 31%. The absolute number is capped
+by the narrative-vs-audiovisual mismatch above — many boundaries have no
+audio-visual signature at all — so the point is the downstream effect, below.
+
+| boundary-F1 @±20 s | grayscale | learned (LOO) | learned (train-all) |
+| ------------------ | --------: | ------------: | ------------------: |
+| macro over 9 films | 29.8% | **44.1%** | 72.9% |
+
+## The result that matters: actor presence
+
+Per-second X-Ray presence F1, macro over the nine films, at the shipped presence
+config. The learned column is **leave-one-out** — each film scored by a detector
+trained on the other eight:
+
+| boundary source for flood-fill | presence F1 |
+| ------------------------------ | ----------- |
+| track-extent (flood off) | 62.6% |
+| flood + grayscale cuts | 64.0% |
+| **flood + learned detector (LOO)** | **74.9%** |
+
+![Macro presence F1 by flood-fill boundary source](assets/images/scene_presence_macro.png)
+
+**+12.3 points over track-extent, +10.9 over the grayscale-cut flood, and it
+improves every one of the nine films — under honest leave-one-out.** Per film:
+
+![Per-film presence F1 by boundary source](assets/images/scene_presence_by_source.png)
+
+| film | track-extent | flood+grayscale | flood+learned (LOO) |
+| ---- | -----------: | --------------: | ------------------: |
+| Benny & Joon | 77.3 | 80.2 | 78.2 |
+| Café Society | 59.1 | 62.2 | 69.8 |
+| Downton Abbey | 41.0 | 51.8 | **78.6** |
+| Lord of War | 74.8 | 77.1 | 77.8 |
+| Lovelace | 70.3 | 74.0 | 78.2 |
+| The Many Saints of Newark | 37.5 | 43.9 | 53.4 |
+| Scarface | 62.6 | **40.9** | **74.9** |
+| Sound of Metal | 75.0 | 78.1 | 86.8 |
+| Valerian | 65.6 | 67.7 | 76.2 |
+
+The two headline cases:
+
+- **Scarface**: the grayscale-cut flood *breaks* it (62.6 → 40.9), because it
+  detects one cut in the whole film. The learned detector — **on a film it never
+  trained on** — takes it to **74.9%**. This is the strongest evidence the
+  detector generalises: it fixes the exact failure that motivated it, held out.
+- **Downton Abbey**: 41.0 (track-extent) → 51.8 (grayscale) → **78.6** — a
+  +37-point swing on the hardest film.
+
+Naive flood-fill barely beat doing nothing (64% vs 62%) and broke a film. With a
+real boundary detector, flood-fill is decisively the right mode.
+
+### What the frames look like
+
+`scripts/optimizer/dump_error_frames.py` pulls representative seconds and draws
+each face box coloured against X-Ray's scene cast: **green** = true positive (a
+name X-Ray also credits to this scene), **red** = false positive (a name X-Ray
+does *not* credit here — the real error), **orange** = an unknown detection. Cast
+X-Ray lists as present but for whom no face was detected — the structural
+false-negatives a face pipeline can never box — are listed as a **blue** panel.
+
+![A correctly identified second: green true-positive boxes](assets/images/scarface_tp_example.jpg)
+
+Above: three faces named correctly (green). Below: the face-vs-scene-cast tension
+made visual — the one visible face is confidently named (here it is a red
+false-positive, a lead X-Ray did not credit to this exact scene), while six
+credited cast members are off-camera with no face to detect (blue). This is why
+recall against X-Ray has a structural ceiling, not a fixable bug.
+
+![A false-positive box (red) with off-screen cast listed (blue)](assets/images/scarface_fn_fp_example.jpg)
+
+## In the pipeline
+
+Boundary detection is a **post-EOF step**, like flood-fill itself: the per-film
+knee needs every peak, so it can only run once the whole film is seen. The
+`camera_position_change_detector` stamps a per-frame RGB histogram onto each frame;
+it rides through to the result sink; at end-of-stream the sink runs the detector
+over the collected histograms plus the movie's audio log-PSD and snaps the
+presence windows to the result. Enable it with:
+
+```bash
+scene_analyze --movie <file> --gallery <gallery.h5> \
+    --scene-xgb-model models/scene_boundary_xgb.json
+```
+
+Inference is real XGBoost, built into the binary via CMake (`SAE_SCENE_XGB`); the
+audio log-PSD uses FFTW + the existing FFmpeg decode. To keep training and
+inference on one feature implementation, the shipped model is **trained on the
+C++-extracted features** (`scene_features_dump` → `train_xgb_cpp.py`) rather than a
+re-implementation in Python — parity by construction. Verified end to end through
+`scene_analyze` on a movie file and through the Jellyfin work-queue worker.
+
+## Reproduce
+
+```bash
+# per-second audio log-PSD for each film
+.venv-rocm/bin/python scripts/scene_detector/extract_audio_features.py \
+    --manifest experiments/manifests/films_LVFace_opencv5.json
+
+# C++ feature matrices (same features training and inference share)
+build/scene_features_dump <dump.h5> <movie> <features.h5>
+
+# train the shipped model on all nine films
+.venv-rocm/bin/python scripts/scene_detector/train_xgb_cpp.py --train-all
+
+# downstream A/B (track-extent vs flood+grayscale vs flood+learned)
+scripts/scene_detector/downstream_presence.py
+```

docs/traceability.md

@@ -3,7 +3,7 @@
 <!-- GENERATED FILE - do not edit by hand. -->
 <!-- Regenerate: scripts/traceability/traceability-gate.sh -->
 
-**Generated:** 2026-08-05T15:50:49+00:00
+**Generated:** 2026-08-08T10:06:24+00:00
 
 Denominators are read from [`requirements.md`](requirements.md) at run time, never hardcoded. Coverage counts a requirement only when it is tagged in source **and** has a verification tier this repo's CI host can execute (`T1, T2, T3, static`).
 
@@ -11,12 +11,12 @@ Denominators are read from [`requirements.md`](requirements.md) at run time, nev
 
 | Metric | Value |
 |---|---|
-| Source files scanned | 118 |
+| Source files scanned | 119 |
-| TRACES tags found | 215 |
+| TRACES tags found | 239 |
 | EXCEPTION tags found | 1 |
-| Requirements defined | 71 |
+| Requirements defined | 72 |
 | Requirements covered | 42 |
-| **Coverage** | **59.2%** (42/71) |
+| **Coverage** | **58.3%** (42/72) |
 | Coverage of CI-executable scope | 73.7% (42/57) |
 | Tagged but unexecuted in CI | 10 |
 | Orphan tags | 0 |
@@ -29,7 +29,7 @@ Denominators are read from [`requirements.md`](requirements.md) at run time, nev
 | DP | 2 | 0 | 8 |
 | IR | 8 | 0 | 8 |
 | GR | 5 | 0 | 9 |
-| VR | 1 | 9 | 16 |
+| VR | 1 | 9 | 17 |
 
 - **UT** tags present (separate taxonomy, not counted in coverage): UT-001, UT-002, UT-003, UT-004, UT-005, UT-101, UT-102, UT-103, UT-104, UT-105, UT-106, UT-107, UT-108, UT-120, UT-121, UT-122, UT-123, UT-124, UT-130, UT-131, UT-132, UT-133, UT-134, UT-135, UT-136, UT-137, UT-138, UT-139, UT-140, UT-141
 - **IT** tags present (separate taxonomy, not counted in coverage): IT-001
@@ -56,6 +56,7 @@ These requirements have no verification tier this repo's CI host can run, so a t
 | VR-012 | T4, out-of-ci | no | Quality-knee study — TPI/FPI vs sharpness and vs pose, as VR-005 did … |
 | VR-013 | T4, out-of-ci | yes | Cross-source identification probe — gallery from one recording, probe… |
 | VR-015 | out-of-ci | yes | Per-node cost and bottleneck attribution for a run — where the time a… |
+| VR-017 | out-of-ci | no | **Vote-lag study** — how often does the matcher fall more than `track… |
 
 **Tagged but unexecuted:** AR-027, VR-001, VR-002, VR-003, VR-004, VR-005, VR-010, VR-011, VR-013, VR-015 — a test exists and is tagged, but this CI host cannot run it. Report those runs separately.
 
@@ -86,19 +87,19 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 | AR-001 | Done | T3 | SR-002 | covered | `src/nodes/face_detector_node.hpp` | Detect faces in sampled frames; emit bbox, confidence, 5-point landma… |
 | AR-002 | **Done** — `FaceDet… | T2 | SR-002 | covered | `src/nodes/face_detector_node.hpp`, `tests/test_face_detector_node.cpp`, `tests/test_replay_fixtures.cpp` | Minimum face size **40×40 px** (VR-013 measured end to end; VR-005's … |
 | AR-003 | **Done** — `max_fac… | T1, T2, T4 | SR-002 | covered | `src/config.hpp`, `src/nodes/face_detector_node.hpp`, `src/nodes/identity_matcher_node.hpp` | No fixed per-frame face cap — crowd scenes must not lose background c… |
-| AR-004 | **Mostly** — node o… | T1, T4 | SR-002 | covered | `src/benchmark.hpp`, `src/main.cpp`, `src/nodes/identity_matcher_node.hpp`, `tests/test_replay_fixtures.cpp` | Backpressure: unbounded faces/frame absorbed by slowing, never by dro… |
+| AR-004 | **Mostly** — node o… | T1, T4 | SR-002 | covered | `scripts/optimizer/replay.py`, `src/benchmark.hpp`, `src/main.cpp`, `src/nodes/identity_matcher_node.hpp`, `src/nodes/scene_detector_node.hpp`, `src/types.hpp`, `tests/test_channel_bytes.cpp`, `tests/test_replay_fixtures.cpp`, `tests/test_scene_detector_node.cpp` | Backpressure: unbounded faces/frame absorbed by slowing, never by dro… |
 | AR-005 | **Done** — `umeyama… | T1, T3 | SR-002 | covered | `src/face_utils.hpp`, `src/nodes/face_aligner_node.hpp`, `tests/test_face_utils.cpp` | Align to 112×112 via ArcFace 5-point similarity transform, fitted by … |
 | AR-006 | Done | T3 | SR-002 | covered | `src/nodes/embedder_node.hpp` | 512-d L2-normalised embeddings, batched |
 | AR-007 | **Done** — `track_a… | T2 | SR-002 | covered | `src/config.hpp`, `src/main.cpp`, `src/nodes/face_tracker_node.hpp`, `src/scene_preview.cpp`, `tests/test_face_tracker.cpp` | Associate detections by IoU + embedding, with **frame-dependent** wei… |
-| AR-008 | **Done** — one pool… | T2 | SR-002 | covered | `src/config.hpp`, `src/main.cpp`, `src/nodes/face_tracker_node.hpp`, `tests/test_face_tracker.cpp` | One track pool keyed on `last_seen`; no separate revival path |
+| AR-008 | **Done** — one pool… | T2 | SR-002 | covered | `src/config.hpp`, `src/main.cpp`, `src/nodes/face_tracker_node.hpp`, `src/track_registry.hpp`, `tests/test_face_tracker.cpp`, `tests/test_track_registry.cpp` | One track pool keyed on `last_seen`; no separate revival path |
 | AR-009 | Done | T2 | SR-002 | covered | `src/nodes/camera_position_change_detector_node.hpp` | Camera-cut detection (histogram) as an association hint |
-| AR-010 | **Done** — decode b… | T2 | SR-002 | covered | `src/main.cpp`, `src/nodes/scene_boundary_annotator_node.hpp`, `src/nodes/scene_detector_node.hpp`, `src/scene_boundaries.hpp` | Scene-boundary detection (TransNetV2) as an association hint |
+| AR-010 | **Done** — decode b… | T2 | SR-002 | covered | `src/main.cpp`, `src/nodes/scene_boundary_annotator_node.hpp`, `src/nodes/scene_detector_node.hpp`, `src/scene_boundaries.hpp`, `tests/test_scene_detector_node.cpp` | Scene-boundary detection (TransNetV2) as an association hint |
 | AR-011 | **Done** — both vio… | T1, T2 | SR-002 | covered | `src/config.hpp`, `src/nodes/scene_detector_node.hpp`, `src/scene_boundaries.hpp`, `tests/test_scene_detector_node.cpp` | **Every model is fed the input it was trained for** — cost reduced by… |
 | AR-012 | **Done** — `src/tra… | T2 | **SR-002** | covered | `src/config.hpp`, `src/kpn_bindings.cpp`, `src/main.cpp`, `src/nodes/frame_annotation_node.hpp`, `src/nodes/identity_matcher_node.hpp`, `src/nodes/result_sink_node.hpp`, `src/scene_preview.cpp`, `src/track_registry.hpp`, `tests/test_replay_fixtures.cpp`, `tests/test_track_registry.cpp` | Presence follows **track extent**, not per-frame recognition |
-| AR-013 | **Done** — `last_se… | T2 | SR-002 | covered | `src/config.hpp`, `src/kpn_bindings.cpp`, `src/nodes/frame_annotation_node.hpp`, `src/track_registry.hpp`, `tests/test_replay_fixtures.cpp`, `tests/test_track_registry.cpp` | `last_seen` optional state machine; window ends at last sighting, nev… |
+| AR-013 | **Done** — `last_se… | T2 | SR-002 | covered | `src/config.hpp`, `src/nodes/frame_annotation_node.hpp`, `src/nodes/identity_matcher_node.hpp`, `src/track_registry.hpp`, `tests/test_replay_fixtures.cpp`, `tests/test_track_registry.cpp` | `last_seen` optional state machine; window ends at last sighting, nev… |
 | AR-014 | **Done** — swap clo… | T2 | SR-002 | covered | `src/track_registry.hpp`, `tests/test_track_registry.cpp` | Belief swap A→B terminates the track and starts a new one |
 | AR-015 | **Done** — reverse … | T2 | SR-002 | covered | `src/track_registry.hpp`, `tests/test_track_registry.cpp` | Two live tracks owned by one actor ⇒ treat as a detected cut, re-asso… |
-| AR-016 | **Done** — `flush()… | T2 | SR-002 | covered | `src/main.cpp`, `src/nodes/result_sink_node.hpp`, `src/track_registry.hpp`, `tests/test_track_registry.cpp` | All tracks closed at EOF — a film ends with faces on screen |
+| AR-016 | **Done** — `flush()… | T2 | SR-002 | covered | `src/kpn_bindings.cpp`, `src/main.cpp`, `src/nodes/result_sink_node.hpp`, `src/track_registry.hpp`, `tests/test_track_registry.cpp` | All tracks closed at EOF — a film ends with faces on screen |
 | AR-017 | **Done** — `DeadTra… | T1, T2 | SR-002 | covered | `src/config.hpp`, `src/nodes/result_sink_node.hpp`, `src/track_registry.hpp`, `tests/test_track_registry.cpp` | Every presence claim carries its belief and identification route |
 | AR-018 | **Done** — banded a… | T1, T2 | SR-005 | covered | `src/config.hpp`, `src/gallery/track_gallery.hpp`, `src/nodes/identity_matcher_node.hpp`, `tests/test_track_gallery.cpp` | Per-subject embedding store with banded admission (novel enough, safe… |
 | AR-019 | **Done** — all thre… | T2 | SR-005 | covered | `src/gallery/track_gallery.hpp`, `src/nodes/identity_matcher_node.hpp`, `tests/test_track_gallery.cpp` | Per-film gallery annex from owned tracks; acquires the non-frontal vi… |
@@ -106,14 +107,14 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 | AR-021 | Planned | T2 | SR-005 | untagged | - | Cluster unknown tracks into one entity per person, under temporal can… |
 | AR-022 | Planned | T1, T2 | §4 | untagged | - | Capture still-unidentified tracks: embeddings, metadata, **context cr… |
 | AR-023 | **Done** — and the … | T1 | SR-002 | covered | `src/gallery/gallery_calibration.hpp`, `src/nodes/identity_matcher_node.hpp`, `tests/test_calibration.cpp` | Fit sigmoid calibration from intra/inter similarity distributions |
-| AR-024 | **Done** — associat… | T1, static | SR-002 | covered | `scripts/ci/check_raw_cosine.py`, `scripts/optimizer/replay.py`, `src/config.hpp`, `src/evidence_discount.hpp`, `src/gallery/gallery_calibration.hpp`, `src/gallery/track_gallery.hpp`, `src/main.cpp`, `src/nodes/face_tracker_node.hpp`, `src/nodes/identity_matcher_node.hpp`, `src/scene_preview.cpp`, `tests/test_track_gallery.cpp` | **Always the calibrated probability, never a raw cosine** — exception… |
+| AR-024 | **Done** — associat… | T1, static | SR-002 | covered | `scripts/ci/check_raw_cosine.py`, `src/config.hpp`, `src/evidence_discount.hpp`, `src/gallery/gallery_calibration.hpp`, `src/gallery/track_gallery.hpp`, `src/main.cpp`, `src/nodes/face_tracker_node.hpp`, `src/nodes/identity_matcher_node.hpp`, `src/scene_preview.cpp`, `tests/test_track_gallery.cpp` | **Always the calibrated probability, never a raw cosine** — exception… |
-| AR-025 | **Done** — log-odds… | T1 | SR-002 | covered | `src/config.hpp`, `src/evidence_discount.hpp`, `src/main.cpp`, `src/nodes/identity_matcher_node.hpp` | Per-track Bayesian accumulation in log-odds, with correlated-observat… |
+| AR-025 | **Done** — log-odds… | T1 | SR-002 | covered | `src/config.hpp`, `src/evidence_discount.hpp`, `src/kpn_bindings.cpp`, `src/main.cpp`, `src/nodes/identity_matcher_node.hpp`, `src/track_registry.hpp`, `tests/test_track_registry.cpp` | Per-track Bayesian accumulation in log-odds, with correlated-observat… |
 | AR-026 | **In Progress** — t… | T1, T4 | SR-001 | covered | `src/backends/gemm_backend.cpp`, `src/gallery/track_gallery.hpp`, `src/inference/similarity.hpp`, `src/nodes/identity_matcher_node.hpp`, `tests/test_similarity.cpp`, `tests/test_track_gallery.cpp` | All similarity computed as GEMM, including annex and deferred pass |
 | AR-027 | Planned | T4 | SR-001 | tagged, unexecuted | `src/backends/gemm_backend.cpp` | Throughput acceptable for **arbitrary** gallery size |
 | AR-028 | **Done** — filled i… | T2 | SR-002 | covered | `scripts/optimizer/replay.py`, `src/nodes/embedding_dump_node.hpp`, `src/nodes/face_aligner_node.hpp`, `src/types.hpp`, `tests/test_embedding_dump.cpp`, `tests/test_face_utils.cpp` | **Embedding input quality assessed and carried** — every face scored … |
 | AR-029 | **Done** — `crop_sh… | T1 | SR-002 | covered | `src/face_utils.hpp`, `src/nodes/face_aligner_node.hpp`, `tests/test_face_utils.cpp` | Sharpness measure on the **aligned crop** (scale-normalised, so it ca… |
 | AR-030 | **In Progress** — m… | T1 | SR-002 | covered | `src/face_utils.hpp`, `src/nodes/face_aligner_node.hpp`, `tests/test_face_utils.cpp` | Visibility measure from the AR-001 5-point landmarks — extreme pose o… |
-| DP-001 | **Done, after a rep… | T1, manual | PR-004 | covered | `src/main.cpp`, `src/scene_preview.cpp` | One analysis core; modes are front-ends and must not fork pipeline lo… |
+| DP-001 | **Done, after a rep… | T1, manual | PR-004 | covered | `scripts/optimizer/replay.py`, `src/kpn_bindings.cpp`, `src/main.cpp`, `src/scene_preview.cpp` | One analysis core; modes are front-ends and must not fork pipeline lo… |
 | DP-002 | Done | T1, manual | PR-004 | covered | `src/main.cpp` | Batch CLI over one title |
 | DP-003 | Planned | T1, manual | PR-004 | untagged | - | On-demand resident service with bounded, observable queue |
 | DP-004 | Planned | T1, manual | PR-004 | untagged | - | Opportunistic/idle mode: external trigger, hard stop, implicit re-que… |
@@ -121,9 +122,9 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 | DP-006 | Planned | T1, manual | PR-003 | untagged | - | Background incremental gallery refresh on a timer |
 | DP-007 | **Mostly** — image … | T1, manual | PR-004 | untagged | - | CI builder image, CPU-only, pinned by tag in the Gitea container regi… |
 | DP-008 | Planned | T1, manual | PR-004 | untagged | - | Builder images + release jobs per backend (cpu / cuda / rocm); ship b… |
-| IR-001 | Done | T1 | SR-003 | covered | `src/nodes/result_sink_node.hpp` | Emit the JRay truth format as sibling `.jray.json` |
+| IR-001 | Done | T1 | SR-003 | covered | `src/kpn_bindings.cpp`, `src/nodes/result_sink_node.hpp` | Emit the JRay truth format as sibling `.jray.json` |
 | IR-002 | **Done** — `schema_… | T1 | SR-003 | covered | `src/config.hpp`, `src/main.cpp`, `src/nodes/result_sink_node.hpp`, `src/track_registry.hpp` | Windows carry belief + route; `extraction.*` carries `extinction_sec`… |
-| IR-003 | **In Progress** — s… | T1 | SR-003 | covered | `src/main.cpp` | Output written **after** the deferred pass, not at EOF |
+| IR-003 | **In Progress** — s… | T1 | SR-003 | covered | `src/kpn_bindings.cpp`, `src/main.cpp` | Output written **after** the deferred pass, not at EOF |
 | IR-004 | **Done** — `src/aud… | T1 | SR-003 | covered | `scripts/validation/test_audio_offset.py`, `src/audio_bindings.cpp`, `src/audio_signature.cpp`, `src/audio_signature.hpp`, `tests/test_audio_signature.cpp` | Compute the audio signature exactly per server spec §3 |
 | IR-005 | **Done** — `tests/f… | T1 | SR-003 | covered | `src/audio_bindings.cpp`, `src/audio_signature.cpp`, `src/audio_signature.hpp`, `tests/test_audio_signature.cpp` | Golden-vector fixture shared with the plugin repo to prove bit-exactn… |
 | IR-006 | Done | T1, manual | SR-001 | covered | `scripts/run_from_jellyfin.py` | Jellyfin round-trip: pull pending queue, push complete results only |
@@ -139,7 +140,7 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 | GR-008 | Planned | T1 | SR-005 | untagged | - | Flag distributional outliers among an actor's references (poisoning g… |
 | GR-009 | TBD | T1 | §4 | untagged | - | Human-confirmed associations persist and improve future extractions |
 | VR-001 | Done | out-of-ci | PR-002 | tagged, unexecuted | `src/nodes/embedding_dump_node.hpp`, `tests/test_embedding_dump.cpp`, `tests/test_replay_fixtures.cpp` | HDF5 post-inference dump at the embedded-frame boundary |
-| VR-002 | **In Progress, and … | out-of-ci | PR-002 | tagged, unexecuted | `scripts/optimizer/replay.py`, `tests/test_replay_fixtures.cpp` | Replay drives the **real** KPN nodes, not a reimplementation |
+| VR-002 | **Done** — includin… | out-of-ci | PR-002 | tagged, unexecuted | `scripts/optimizer/replay.py`, `src/kpn_bindings.cpp`, `tests/test_replay_fixtures.cpp` | Replay drives the **real** KPN nodes, not a reimplementation |
 | VR-003 | Done | out-of-ci | PR-002 | tagged, unexecuted | `scripts/optimizer/second_score.py` | Scoring: micro-F1 against X-Ray, precision/recall logged at every eva… |
 | VR-004 | Done | out-of-ci | PR-002 | tagged, unexecuted | `scripts/validation/ground_truth.py` | Reproducible validation corpus with ground truth |
 | VR-005 | **Done** — knee at … | out-of-ci | PR-002 | tagged, unexecuted | `scripts/validation/min_face_size.py` | Minimum face size study — TPI/FPI vs probe size, gallery held at nati… |
@@ -148,12 +149,13 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 | VR-008 | Planned | out-of-ci | PR-002 | untagged | - | Gallery scaling benchmark — throughput vs gallery size |
 | VR-009 | Planned | T1, out-of-ci | PR-002 | untagged | - | Verify accumulated posteriors are calibrated against held-out tracks |
 | VR-010 | **Done** — `DumpPro… | out-of-ci | PR-002 | tagged, unexecuted | `src/nodes/embedding_dump_node.hpp` | Dump provenance attributes — embedder model, detector settings, `dens… |
-| VR-011 | Planned | out-of-ci | PR-002 | tagged, unexecuted | `scripts/optimizer/replay.py` | Rewrite the replay harness for the post-AR-012 output contract |
+| VR-011 | **Done** — `sae_kpn… | out-of-ci | PR-002 | tagged, unexecuted | `scripts/optimizer/replay.py`, `scripts/optimizer/test_sae_kpn.py`, `src/kpn_bindings.cpp` | Rewrite the replay harness for the post-AR-012 output contract |
 | VR-012 | Planned | T4, out-of-ci | PR-002 | untagged | - | Quality-knee study — TPI/FPI vs sharpness and vs pose, as VR-005 did … |
 | VR-013 | **In Progress** — h… | T4, out-of-ci | PR-002 | tagged, unexecuted | `experiments/xsource/resolution_sweep.py`, `experiments/xsource/verify_labels.py` | Cross-source identification probe — gallery from one recording, probe… |
 | VR-014 | **Done** — 40 rando… | T2, out-of-ci | PR-002 | covered | `scripts/validation/test_audio_offset.py` | Audio-signature **offset recovery on real content** — a known trim re… |
 | VR-015 | **Done** — `--bench… | out-of-ci | PR-004 | tagged, unexecuted | `src/backends/trt_backend.cpp`, `src/benchmark.hpp`, `src/config.hpp`, `src/main.cpp`, `tests/test_benchmark.cpp` | Per-node cost and bottleneck attribution for a run — where the time a… |
 | VR-016 | **Planned.** The hi… | T2, out-of-ci | PR-002 | untagged | - | **Cut-detection cadence study** — is `cut_threshold` 0.70 meaningful … |
+| VR-017 | **Planned.** Channe… | out-of-ci | PR-002 | untagged | - | **Vote-lag study** — how often does the matcher fall more than `track… |
 
 ## Detailed mapping
 
@@ -177,21 +179,29 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 - [`src/config.hpp:52`](../src/config.hpp#L52) — `Unknown`
 - [`src/nodes/face_detector_node.hpp:64`](../src/nodes/face_detector_node.hpp#L64) — `private:`
-- [`src/nodes/identity_matcher_node.hpp:192`](../src/nodes/identity_matcher_node.hpp#L192) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
+- [`src/nodes/identity_matcher_node.hpp:213`](../src/nodes/identity_matcher_node.hpp#L213) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
 
 ### AR-004
 
-**Locations:** 9
+**Locations:** 17
 
 - [`src/benchmark.hpp:176`](../src/benchmark.hpp#L176) — `Unknown`
 - [`src/benchmark.hpp:472`](../src/benchmark.hpp#L472) — `void print(std::ostream& os, double film_sec) const`
 - [`src/benchmark.hpp:522`](../src/benchmark.hpp#L522) — `else if (c.in_fill_pct > 50.0)`
-- [`src/main.cpp:103`](../src/main.cpp#L103) — `static constexpr std::size_t kSceneJoinDepth = 256;`
+- [`src/main.cpp:106`](../src/main.cpp#L106) — `static constexpr std::size_t kSceneInputDepth = 128;`
-- [`src/main.cpp:115`](../src/main.cpp#L115) — `static std::shared_ptr<SceneBoundaries> scene_stats;`
+- [`src/main.cpp:111`](../src/main.cpp#L111) — `static constexpr std::size_t kSceneInputDepth = 128;`
-- [`src/main.cpp:394`](../src/main.cpp#L394) — `Unknown`
+- [`src/main.cpp:132`](../src/main.cpp#L132) — `static constexpr double kSceneJoinSafety = 2.0;`
-- [`src/main.cpp:437`](../src/main.cpp#L437) — `std::ofstream bf(cfg.benchmark_path);`
+- [`src/main.cpp:154`](../src/main.cpp#L154) — `static std::size_t scene_join_depth(float sample_fps)`
-- [`src/nodes/identity_matcher_node.hpp:192`](../src/nodes/identity_matcher_node.hpp#L192) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
+- [`src/main.cpp:172`](../src/main.cpp#L172) — `static std::shared_ptr<SceneBoundaries> scene_stats;`
+- [`src/main.cpp:451`](../src/main.cpp#L451) — `Unknown`
+- [`src/main.cpp:494`](../src/main.cpp#L494) — `std::ofstream bf(cfg.benchmark_path);`
+- [`src/nodes/identity_matcher_node.hpp:213`](../src/nodes/identity_matcher_node.hpp#L213) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
+- [`src/nodes/scene_detector_node.hpp:95`](../src/nodes/scene_detector_node.hpp#L95) — `Unknown`
+- [`src/types.hpp:189`](../src/types.hpp#L189) — `Unknown`
+- [`tests/test_channel_bytes.cpp:3`](../tests/test_channel_bytes.cpp#L3) — `Unknown`
 - [`tests/test_replay_fixtures.cpp:3`](../tests/test_replay_fixtures.cpp#L3) — `Unknown`
+- [`tests/test_scene_detector_node.cpp:93`](../tests/test_scene_detector_node.cpp#L93) — `Unknown`
+- [`scripts/optimizer/replay.py:174`](../scripts/optimizer/replay.py#L174) — `if i < len(frames):`
 
 ### AR-005
 
@@ -212,19 +222,21 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 **Locations:** 5
 
 - [`src/config.hpp:160`](../src/config.hpp#L160) — `Unknown`
-- [`src/main.cpp:260`](../src/main.cpp#L260) — `Unknown`
+- [`src/main.cpp:317`](../src/main.cpp#L317) — `Unknown`
 - [`src/nodes/face_tracker_node.hpp:2`](../src/nodes/face_tracker_node.hpp#L2) — `Unknown`
 - [`src/scene_preview.cpp:147`](../src/scene_preview.cpp#L147) — `Unknown`
 - [`tests/test_face_tracker.cpp:1`](../tests/test_face_tracker.cpp#L1) — `Unknown`
 
 ### AR-008
 
-**Locations:** 4
+**Locations:** 6
 
 - [`src/config.hpp:160`](../src/config.hpp#L160) — `Unknown`
-- [`src/main.cpp:260`](../src/main.cpp#L260) — `Unknown`
+- [`src/main.cpp:317`](../src/main.cpp#L317) — `Unknown`
 - [`src/nodes/face_tracker_node.hpp:2`](../src/nodes/face_tracker_node.hpp#L2) — `Unknown`
+- [`src/track_registry.hpp:145`](../src/track_registry.hpp#L145) — `public:`
 - [`tests/test_face_tracker.cpp:1`](../tests/test_face_tracker.cpp#L1) — `Unknown`
+- [`tests/test_track_registry.cpp:417`](../tests/test_track_registry.cpp#L417) — `TrackRegistry reg(cfg(/*extinction=*/5.0), disc());`
 
 ### AR-009
 
@@ -234,57 +246,69 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### AR-010
 
-**Locations:** 9
+**Locations:** 12
 
-- [`src/main.cpp:103`](../src/main.cpp#L103) — `static constexpr std::size_t kSceneJoinDepth = 256;`
+- [`src/main.cpp:106`](../src/main.cpp#L106) — `static constexpr std::size_t kSceneInputDepth = 128;`
-- [`src/main.cpp:447`](../src/main.cpp#L447) — `Unknown`
+- [`src/main.cpp:111`](../src/main.cpp#L111) — `static constexpr std::size_t kSceneInputDepth = 128;`
-- [`src/main.cpp:517`](../src/main.cpp#L517) — `return run_net(std::move(net));`
+- [`src/main.cpp:504`](../src/main.cpp#L504) — `Unknown`
-- [`src/main.cpp:549`](../src/main.cpp#L549) — `Unknown`
+- [`src/main.cpp:613`](../src/main.cpp#L613) — `return run_net(std::move(net));`
+- [`src/main.cpp:645`](../src/main.cpp#L645) — `Unknown`
 - [`src/nodes/scene_boundary_annotator_node.hpp:2`](../src/nodes/scene_boundary_annotator_node.hpp#L2) — `Unknown`
-- [`src/nodes/scene_detector_node.hpp:36`](../src/nodes/scene_detector_node.hpp#L36) — `static constexpr std::string_view label() { return "scene_detector"; }`
+- [`src/nodes/scene_detector_node.hpp:38`](../src/nodes/scene_detector_node.hpp#L38) — `static constexpr std::string_view label() { return "scene_detector"; }`
-- [`src/nodes/scene_detector_node.hpp:147`](../src/nodes/scene_detector_node.hpp#L147) — `Unknown`
+- [`src/nodes/scene_detector_node.hpp:95`](../src/nodes/scene_detector_node.hpp#L95) — `Unknown`
-- [`src/nodes/scene_detector_node.hpp:188`](../src/nodes/scene_detector_node.hpp#L188) — `void write_output()`
+- [`src/nodes/scene_detector_node.hpp:191`](../src/nodes/scene_detector_node.hpp#L191) — `Unknown`
+- [`src/nodes/scene_detector_node.hpp:232`](../src/nodes/scene_detector_node.hpp#L232) — `void write_output()`
 - [`src/scene_boundaries.hpp:2`](../src/scene_boundaries.hpp#L2) — `Unknown`
+- [`tests/test_scene_detector_node.cpp:93`](../tests/test_scene_detector_node.cpp#L93) — `Unknown`
 
 ### AR-011
 
 **Locations:** 6
 
 - [`src/config.hpp:138`](../src/config.hpp#L138) — `Unknown`
-- [`src/nodes/scene_detector_node.hpp:70`](../src/nodes/scene_detector_node.hpp#L70) — `void operator()(Frame f)`
+- [`src/nodes/scene_detector_node.hpp:72`](../src/nodes/scene_detector_node.hpp#L72) — `void operator()(Frame f)`
-- [`src/nodes/scene_detector_node.hpp:93`](../src/nodes/scene_detector_node.hpp#L93) — `Unknown`
+- [`src/nodes/scene_detector_node.hpp:137`](../src/nodes/scene_detector_node.hpp#L137) — `Unknown`
-- [`src/nodes/scene_detector_node.hpp:152`](../src/nodes/scene_detector_node.hpp#L152) — `Unknown`
+- [`src/nodes/scene_detector_node.hpp:196`](../src/nodes/scene_detector_node.hpp#L196) — `Unknown`
 - [`src/scene_boundaries.hpp:30`](../src/scene_boundaries.hpp#L30) — `public:`
 - [`tests/test_scene_detector_node.cpp:4`](../tests/test_scene_detector_node.cpp#L4) — `Unknown`
 
 ### AR-012
 
-**Locations:** 13
+**Locations:** 17
 
 - [`src/config.hpp:210`](../src/config.hpp#L210) — `Unknown`
-- [`src/kpn_bindings.cpp:273`](../src/kpn_bindings.cpp#L273) — `Unknown`
+- [`src/kpn_bindings.cpp:406`](../src/kpn_bindings.cpp#L406) — `Unknown`
-- [`src/main.cpp:260`](../src/main.cpp#L260) — `Unknown`
+- [`src/main.cpp:317`](../src/main.cpp#L317) — `Unknown`
-- [`src/main.cpp:286`](../src/main.cpp#L286) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
+- [`src/main.cpp:343`](../src/main.cpp#L343) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
+- [`src/main.cpp:519`](../src/main.cpp#L519) — `Unknown`
 - [`src/nodes/frame_annotation_node.hpp:2`](../src/nodes/frame_annotation_node.hpp#L2) — `Unknown`
 - [`src/nodes/identity_matcher_node.hpp:151`](../src/nodes/identity_matcher_node.hpp#L151) — `const GalleryCalibration& calibration() const { return cal_; }`
-- [`src/nodes/identity_matcher_node.hpp:285`](../src/nodes/identity_matcher_node.hpp#L285) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:173`](../src/nodes/identity_matcher_node.hpp#L173) — `MatchedSceneFrame operator()(TrackedSceneFrame tf)`
+- [`src/nodes/identity_matcher_node.hpp:306`](../src/nodes/identity_matcher_node.hpp#L306) — `Unknown`
 - [`src/nodes/result_sink_node.hpp:50`](../src/nodes/result_sink_node.hpp#L50) — `static constexpr std::string_view label() { return "result_sink"; }`
 - [`src/nodes/result_sink_node.hpp:164`](../src/nodes/result_sink_node.hpp#L164) — `struct ActorMeta { std::string name, imdb_id, tmdb_id, jellyfin_id; };`
 - [`src/scene_preview.cpp:147`](../src/scene_preview.cpp#L147) — `Unknown`
 - [`src/track_registry.hpp:2`](../src/track_registry.hpp#L2) — `Unknown`
+- [`src/track_registry.hpp:191`](../src/track_registry.hpp#L191) — `void tick(double now) { std::lock_guard g(mu_); tick_locked(now); }`
 - [`tests/test_replay_fixtures.cpp:3`](../tests/test_replay_fixtures.cpp#L3) — `Unknown`
 - [`tests/test_track_registry.cpp:3`](../tests/test_track_registry.cpp#L3) — `Unknown`
+- [`tests/test_track_registry.cpp:376`](../tests/test_track_registry.cpp#L376) — `TrackRegistry reg(cfg(/*extinction=*/5.0), disc());`
 
 ### AR-013
 
-**Locations:** 6
+**Locations:** 11
 
 - [`src/config.hpp:210`](../src/config.hpp#L210) — `Unknown`
-- [`src/kpn_bindings.cpp:273`](../src/kpn_bindings.cpp#L273) — `Unknown`
 - [`src/nodes/frame_annotation_node.hpp:2`](../src/nodes/frame_annotation_node.hpp#L2) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:173`](../src/nodes/identity_matcher_node.hpp#L173) — `MatchedSceneFrame operator()(TrackedSceneFrame tf)`
 - [`src/track_registry.hpp:2`](../src/track_registry.hpp#L2) — `Unknown`
+- [`src/track_registry.hpp:145`](../src/track_registry.hpp#L145) — `public:`
+- [`src/track_registry.hpp:191`](../src/track_registry.hpp#L191) — `void tick(double now) { std::lock_guard g(mu_); tick_locked(now); }`
+- [`src/track_registry.hpp:222`](../src/track_registry.hpp#L222) — `std::lock_guard g(mu_);`
 - [`tests/test_replay_fixtures.cpp:3`](../tests/test_replay_fixtures.cpp#L3) — `Unknown`
 - [`tests/test_track_registry.cpp:3`](../tests/test_track_registry.cpp#L3) — `Unknown`
+- [`tests/test_track_registry.cpp:376`](../tests/test_track_registry.cpp#L376) — `TrackRegistry reg(cfg(/*extinction=*/5.0), disc());`
+- [`tests/test_track_registry.cpp:417`](../tests/test_track_registry.cpp#L417) — `TrackRegistry reg(cfg(/*extinction=*/5.0), disc());`
 
 ### AR-014
 
@@ -302,9 +326,10 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### AR-016
 
-**Locations:** 4
+**Locations:** 5
 
-- [`src/main.cpp:286`](../src/main.cpp#L286) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
+- [`src/kpn_bindings.cpp:406`](../src/kpn_bindings.cpp#L406) — `Unknown`
+- [`src/main.cpp:343`](../src/main.cpp#L343) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
 - [`src/nodes/result_sink_node.hpp:64`](../src/nodes/result_sink_node.hpp#L64) — `void set_pre_write_hook(std::function<void(double)> fn) { pre_write_ = std::move(fn); }`
 - [`src/track_registry.hpp:2`](../src/track_registry.hpp#L2) — `Unknown`
 - [`tests/test_track_registry.cpp:3`](../tests/test_track_registry.cpp#L3) — `Unknown`
@@ -336,9 +361,9 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/gallery/track_gallery.hpp:136`](../src/gallery/track_gallery.hpp#L136) — `Unknown`
 - [`src/gallery/track_gallery.hpp:161`](../src/gallery/track_gallery.hpp#L161) — `Unknown`
 - [`src/gallery/track_gallery.hpp:183`](../src/gallery/track_gallery.hpp#L183) — `void set_owner(int track_id, int actor_idx)`
-- [`src/nodes/identity_matcher_node.hpp:173`](../src/nodes/identity_matcher_node.hpp#L173) — `MatchedSceneFrame operator()(TrackedSceneFrame tf)`
+- [`src/nodes/identity_matcher_node.hpp:194`](../src/nodes/identity_matcher_node.hpp#L194) — `Unknown`
-- [`src/nodes/identity_matcher_node.hpp:295`](../src/nodes/identity_matcher_node.hpp#L295) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:316`](../src/nodes/identity_matcher_node.hpp#L316) — `Unknown`
-- [`src/nodes/identity_matcher_node.hpp:314`](../src/nodes/identity_matcher_node.hpp#L314) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:335`](../src/nodes/identity_matcher_node.hpp#L335) — `Unknown`
 - [`tests/test_track_gallery.cpp:1`](../tests/test_track_gallery.cpp#L1) — `Unknown`
 
 ### AR-023
@@ -352,7 +377,7 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### AR-024
 
-**Locations:** 20
+**Locations:** 19
 
 - [`src/config.hpp:108`](../src/config.hpp#L108) — `Unknown`
 - [`src/config.hpp:160`](../src/config.hpp#L160) — `Unknown`
@@ -362,28 +387,32 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/gallery/track_gallery.hpp:193`](../src/gallery/track_gallery.hpp#L193) — `void set_owner(int track_id, int actor_idx)`
 - [`src/gallery/track_gallery.hpp:233`](../src/gallery/track_gallery.hpp#L233) — `struct TrackState`
 - [`src/gallery/track_gallery.hpp:331`](../src/gallery/track_gallery.hpp#L331) — `Unknown`
-- [`src/main.cpp:260`](../src/main.cpp#L260) — `Unknown`
+- [`src/main.cpp:317`](../src/main.cpp#L317) — `Unknown`
 - [`src/nodes/face_tracker_node.hpp:2`](../src/nodes/face_tracker_node.hpp#L2) — `Unknown`
 - [`src/nodes/identity_matcher_node.hpp:29`](../src/nodes/identity_matcher_node.hpp#L29) — `Unknown`
 - [`src/nodes/identity_matcher_node.hpp:109`](../src/nodes/identity_matcher_node.hpp#L109) — `Unknown`
 - [`src/nodes/identity_matcher_node.hpp:136`](../src/nodes/identity_matcher_node.hpp#L136) — `std::vector<float> host_gallery(static_cast<size_t>(n_gallery_) * 512);`
 - [`src/nodes/identity_matcher_node.hpp:143`](../src/nodes/identity_matcher_node.hpp#L143) — `const GalleryCalibration& calibration() const { return cal_; }`
-- [`src/nodes/identity_matcher_node.hpp:250`](../src/nodes/identity_matcher_node.hpp#L250) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:271`](../src/nodes/identity_matcher_node.hpp#L271) — `Unknown`
 - [`src/scene_preview.cpp:147`](../src/scene_preview.cpp#L147) — `Unknown`
 - [`tests/test_track_gallery.cpp:1`](../tests/test_track_gallery.cpp#L1) — `Unknown`
 - [`tests/test_track_gallery.cpp:414`](../tests/test_track_gallery.cpp#L414) — `TrackGallery tg(expand_cfg());`
 - [`scripts/ci/check_raw_cosine.py:4`](../scripts/ci/check_raw_cosine.py#L4) — `Unknown`
-- [`scripts/optimizer/replay.py:262`](../scripts/optimizer/replay.py#L262) — `Unknown`
 
 ### AR-025
 
-**Locations:** 5
+**Locations:** 10
 
 - [`src/config.hpp:181`](../src/config.hpp#L181) — `Unknown`
 - [`src/evidence_discount.hpp:2`](../src/evidence_discount.hpp#L2) — `Unknown`
-- [`src/main.cpp:268`](../src/main.cpp#L268) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
+- [`src/kpn_bindings.cpp:432`](../src/kpn_bindings.cpp#L432) — `Unknown`
+- [`src/main.cpp:325`](../src/main.cpp#L325) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
+- [`src/main.cpp:519`](../src/main.cpp#L519) — `Unknown`
 - [`src/nodes/identity_matcher_node.hpp:151`](../src/nodes/identity_matcher_node.hpp#L151) — `const GalleryCalibration& calibration() const { return cal_; }`
-- [`src/nodes/identity_matcher_node.hpp:285`](../src/nodes/identity_matcher_node.hpp#L285) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:306`](../src/nodes/identity_matcher_node.hpp#L306) — `Unknown`
+- [`src/track_registry.hpp:191`](../src/track_registry.hpp#L191) — `void tick(double now) { std::lock_guard g(mu_); tick_locked(now); }`
+- [`src/track_registry.hpp:222`](../src/track_registry.hpp#L222) — `std::lock_guard g(mu_);`
+- [`tests/test_track_registry.cpp:376`](../tests/test_track_registry.cpp#L376) — `TrackRegistry reg(cfg(/*extinction=*/5.0), disc());`
 
 ### AR-026
 
@@ -399,8 +428,8 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/inference/similarity.hpp:19`](../src/inference/similarity.hpp#L19) — `struct ISimilarityEngine`
 - [`src/inference/similarity.hpp:39`](../src/inference/similarity.hpp#L39) — `virtual int n_gallery() const = 0;`
 - [`src/nodes/identity_matcher_node.hpp:60`](../src/nodes/identity_matcher_node.hpp#L60) — `struct IdentityMatcherFunc`
-- [`src/nodes/identity_matcher_node.hpp:212`](../src/nodes/identity_matcher_node.hpp#L212) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
+- [`src/nodes/identity_matcher_node.hpp:233`](../src/nodes/identity_matcher_node.hpp#L233) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
-- [`src/nodes/identity_matcher_node.hpp:328`](../src/nodes/identity_matcher_node.hpp#L328) — `private:`
+- [`src/nodes/identity_matcher_node.hpp:349`](../src/nodes/identity_matcher_node.hpp#L349) — `private:`
 - [`tests/test_similarity.cpp:1`](../tests/test_similarity.cpp#L1) — `Unknown`
 - [`tests/test_similarity.cpp:85`](../tests/test_similarity.cpp#L85) — `Unknown`
 - [`tests/test_track_gallery.cpp:1`](../tests/test_track_gallery.cpp#L1) — `Unknown`
@@ -424,7 +453,7 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/types.hpp:62`](../src/types.hpp#L62) — `struct DetectedFace`
 - [`tests/test_embedding_dump.cpp:1`](../tests/test_embedding_dump.cpp#L1) — `Unknown`
 - [`tests/test_face_utils.cpp:1`](../tests/test_face_utils.cpp#L1) — `Unknown`
-- [`scripts/optimizer/replay.py:66`](../scripts/optimizer/replay.py#L66) — `for i in range(len(ts)):`
+- [`scripts/optimizer/replay.py:70`](../scripts/optimizer/replay.py#L70) — `for i in range(len(ts)):`
 
 ### AR-029
 
@@ -445,10 +474,12 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### DP-001
 
-**Locations:** 2
+**Locations:** 4
 
+- [`src/kpn_bindings.cpp:322`](../src/kpn_bindings.cpp#L322) — `Unknown`
 - [`src/main.cpp:3`](../src/main.cpp#L3) — `Unknown`
 - [`src/scene_preview.cpp:147`](../src/scene_preview.cpp#L147) — `Unknown`
+- [`scripts/optimizer/replay.py:194`](../scripts/optimizer/replay.py#L194) — `Unknown`
 
 ### DP-002
 
@@ -500,9 +531,9 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/gallery/gallery_store.cpp:82`](../src/gallery/gallery_store.cpp#L82) — `H5::StrType str(H5::PredType::C_S1, H5T_VARIABLE);`
 - [`src/gallery/gallery_store.cpp:167`](../src/gallery/gallery_store.cpp#L167) — `H5::DataSpace scalar(H5S_SCALAR);`
 - [`src/gallery/gallery_store.cpp:219`](../src/gallery/gallery_store.cpp#L219) — `Unknown`
-- [`src/kpn_bindings.cpp:183`](../src/kpn_bindings.cpp#L183) — `Unknown`
+- [`src/kpn_bindings.cpp:257`](../src/kpn_bindings.cpp#L257) — `Unknown`
-- [`src/kpn_bindings.cpp:233`](../src/kpn_bindings.cpp#L233) — `Unknown`
+- [`src/kpn_bindings.cpp:359`](../src/kpn_bindings.cpp#L359) — `static std::map<std::string, std::shared_ptr<ActorGallery>> cache;`
-- [`src/main.cpp:235`](../src/main.cpp#L235) — `Unknown`
+- [`src/main.cpp:292`](../src/main.cpp#L292) — `Unknown`
 - [`src/nodes/embedding_dump_node.hpp:127`](../src/nodes/embedding_dump_node.hpp#L127) — `static constexpr std::string_view label() { return "embedding_dump"; }`
 - [`src/nodes/embedding_dump_node.hpp:257`](../src/nodes/embedding_dump_node.hpp#L257) — `H5::H5File file(path_, H5F_ACC_TRUNC);`
 - [`src/scene_preview.cpp:130`](../src/scene_preview.cpp#L130) — `int main(int argc, char** argv)`
@@ -530,8 +561,8 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`scripts/optimizer/optimize.py:186`](../scripts/optimizer/optimize.py#L186) — `Unknown`
 - [`scripts/optimizer/optimize.py:202`](../scripts/optimizer/optimize.py#L202) — `if not Path(f["dump"]).exists():`
 - [`scripts/optimizer/reembed_gallery.py:62`](../scripts/optimizer/reembed_gallery.py#L62) — `for i, a in enumerate(ref["actors"], 1):`
-- [`scripts/optimizer/replay.py:125`](../scripts/optimizer/replay.py#L125) — `Unknown`
+- [`scripts/optimizer/replay.py:144`](../scripts/optimizer/replay.py#L144) — `Unknown`
-- [`scripts/optimizer/replay.py:306`](../scripts/optimizer/replay.py#L306) — `Unknown`
+- [`scripts/optimizer/replay.py:337`](../scripts/optimizer/replay.py#L337) — `Unknown`
 - [`scripts/sae_embed_loader.py:23`](../scripts/sae_embed_loader.py#L23) — `def resolve_arcface(models_dir: str, arcface: str \| None = None) -> str:`
 - [`scripts/sae_gallery.py:171`](../scripts/sae_gallery.py#L171) — `if not _stamp_empty(embedder):`
 - [`scripts/sae_gallery.py:200`](../scripts/sae_gallery.py#L200) — `for a in range(len(offset)):`
@@ -546,8 +577,9 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### IR-001
 
-**Locations:** 1
+**Locations:** 2
 
+- [`src/kpn_bindings.cpp:261`](../src/kpn_bindings.cpp#L261) — `Unknown`
 - [`src/nodes/result_sink_node.hpp:2`](../src/nodes/result_sink_node.hpp#L2) — `Unknown`
 
 ### IR-002
@@ -555,7 +587,7 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 **Locations:** 6
 
 - [`src/config.hpp:20`](../src/config.hpp#L20) — `struct Config`
-- [`src/main.cpp:286`](../src/main.cpp#L286) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
+- [`src/main.cpp:343`](../src/main.cpp#L343) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
 - [`src/nodes/result_sink_node.hpp:50`](../src/nodes/result_sink_node.hpp#L50) — `static constexpr std::string_view label() { return "result_sink"; }`
 - [`src/nodes/result_sink_node.hpp:123`](../src/nodes/result_sink_node.hpp#L123) — `void write_output()`
 - [`src/nodes/result_sink_node.hpp:164`](../src/nodes/result_sink_node.hpp#L164) — `struct ActorMeta { std::string name, imdb_id, tmdb_id, jellyfin_id; };`
@@ -563,9 +595,10 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### IR-003
 
-**Locations:** 1
+**Locations:** 2
 
-- [`src/main.cpp:286`](../src/main.cpp#L286) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
+- [`src/kpn_bindings.cpp:406`](../src/kpn_bindings.cpp#L406) — `Unknown`
+- [`src/main.cpp:343`](../src/main.cpp#L343) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
 
 ### IR-004
 
@@ -640,23 +673,34 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### PR-002
 
-**Locations:** 11
+**Locations:** 22
 
+- [`src/kpn_bindings.cpp:6`](../src/kpn_bindings.cpp#L6) — `Unknown`
+- [`src/kpn_bindings.cpp:70`](../src/kpn_bindings.cpp#L70) — `namespace nb = nanobind;`
+- [`src/kpn_bindings.cpp:261`](../src/kpn_bindings.cpp#L261) — `Unknown`
+- [`src/kpn_bindings.cpp:322`](../src/kpn_bindings.cpp#L322) — `Unknown`
+- [`src/kpn_bindings.cpp:432`](../src/kpn_bindings.cpp#L432) — `Unknown`
 - [`src/nodes/embedding_dump_node.hpp:2`](../src/nodes/embedding_dump_node.hpp#L2) — `Unknown`
 - [`src/nodes/embedding_dump_node.hpp:18`](../src/nodes/embedding_dump_node.hpp#L18) — `Unknown`
 - [`src/nodes/embedding_dump_node.hpp:133`](../src/nodes/embedding_dump_node.hpp#L133) — `static constexpr std::string_view label() { return "embedding_dump"; }`
 - [`src/nodes/embedding_dump_node.hpp:159`](../src/nodes/embedding_dump_node.hpp#L159) — `void operator()(EmbeddedSceneFrame ef)`
 - [`src/nodes/embedding_dump_node.hpp:261`](../src/nodes/embedding_dump_node.hpp#L261) — `H5::H5File file(path_, H5F_ACC_TRUNC);`
 - [`scripts/optimizer/replay.py:5`](../scripts/optimizer/replay.py#L5) — `Reads an embedding dump (scripts/optimizer/SCHEMA.md), feeds each frame as an`
-- [`scripts/optimizer/replay.py:212`](../scripts/optimizer/replay.py#L212) — `def build_minimal(annotations, movie, fps, cfg) -> dict:`
+- [`scripts/optimizer/replay.py:120`](../scripts/optimizer/replay.py#L120) — `ending at the last sighting (AR-013) -- and are byte-for-byte the same`
+- [`scripts/optimizer/replay.py:174`](../scripts/optimizer/replay.py#L174) — `if i < len(frames):`
+- [`scripts/optimizer/replay.py:194`](../scripts/optimizer/replay.py#L194) — `Unknown`
+- [`scripts/optimizer/replay.py:242`](../scripts/optimizer/replay.py#L242) — `Unknown`
+- [`scripts/optimizer/replay.py:268`](../scripts/optimizer/replay.py#L268) — `def write_raw_frames(truth: dict, raw_out: str) -> None:`
+- [`scripts/optimizer/replay.py:316`](../scripts/optimizer/replay.py#L316) — `def main():`
 - [`scripts/optimizer/second_score.py:5`](../scripts/optimizer/second_score.py#L5) — `Unknown`
+- [`scripts/optimizer/test_sae_kpn.py:7`](../scripts/optimizer/test_sae_kpn.py#L7) — `Unknown`
 - [`scripts/validation/ground_truth.py:24`](../scripts/validation/ground_truth.py#L24) — `Unknown`
 - [`experiments/xsource/resolution_sweep.py:4`](../experiments/xsource/resolution_sweep.py#L4) — `Unknown`
 - [`experiments/xsource/verify_labels.py:4`](../experiments/xsource/verify_labels.py#L4) — `Unknown`
 
 ### PR-004
 
-**Locations:** 23
+**Locations:** 24
 
 - [`src/backends/trt_backend.cpp:49`](../src/backends/trt_backend.cpp#L49) — `throw CudaError(std::string(what) + ": " + cudaGetErrorString(e));`
 - [`src/benchmark.hpp:2`](../src/benchmark.hpp#L2) — `Unknown`
@@ -672,13 +716,14 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/benchmark.hpp:472`](../src/benchmark.hpp#L472) — `void print(std::ostream& os, double film_sec) const`
 - [`src/benchmark.hpp:522`](../src/benchmark.hpp#L522) — `else if (c.in_fill_pct > 50.0)`
 - [`src/config.hpp:35`](../src/config.hpp#L35) — `Unknown`
+- [`src/kpn_bindings.cpp:322`](../src/kpn_bindings.cpp#L322) — `Unknown`
 - [`src/main.cpp:3`](../src/main.cpp#L3) — `Unknown`
-- [`src/main.cpp:115`](../src/main.cpp#L115) — `static std::shared_ptr<SceneBoundaries> scene_stats;`
+- [`src/main.cpp:172`](../src/main.cpp#L172) — `static std::shared_ptr<SceneBoundaries> scene_stats;`
-- [`src/main.cpp:208`](../src/main.cpp#L208) — `int main(int argc, char** argv)`
+- [`src/main.cpp:265`](../src/main.cpp#L265) — `int main(int argc, char** argv)`
-- [`src/main.cpp:294`](../src/main.cpp#L294) — `Unknown`
+- [`src/main.cpp:351`](../src/main.cpp#L351) — `Unknown`
-- [`src/main.cpp:368`](../src/main.cpp#L368) — `std::lock_guard<std::mutex> lk(event_mtx);`
+- [`src/main.cpp:425`](../src/main.cpp#L425) — `std::lock_guard<std::mutex> lk(event_mtx);`
-- [`src/main.cpp:394`](../src/main.cpp#L394) — `Unknown`
+- [`src/main.cpp:451`](../src/main.cpp#L451) — `Unknown`
-- [`src/main.cpp:407`](../src/main.cpp#L407) — `Unknown`
+- [`src/main.cpp:464`](../src/main.cpp#L464) — `Unknown`
 - [`src/scene_preview.cpp:147`](../src/scene_preview.cpp#L147) — `Unknown`
 - [`tests/test_benchmark.cpp:3`](../tests/test_benchmark.cpp#L3) — `Unknown`
 
@@ -715,14 +760,14 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/gallery/track_gallery.hpp:380`](../src/gallery/track_gallery.hpp#L380) — `static constexpr int kEmbDim = 512;`
 - [`src/inference/similarity.hpp:19`](../src/inference/similarity.hpp#L19) — `struct ISimilarityEngine`
 - [`src/inference/similarity.hpp:39`](../src/inference/similarity.hpp#L39) — `virtual int n_gallery() const = 0;`
-- [`src/kpn_bindings.cpp:183`](../src/kpn_bindings.cpp#L183) — `Unknown`
+- [`src/kpn_bindings.cpp:257`](../src/kpn_bindings.cpp#L257) — `Unknown`
-- [`src/kpn_bindings.cpp:233`](../src/kpn_bindings.cpp#L233) — `Unknown`
+- [`src/kpn_bindings.cpp:359`](../src/kpn_bindings.cpp#L359) — `static std::map<std::string, std::shared_ptr<ActorGallery>> cache;`
-- [`src/main.cpp:235`](../src/main.cpp#L235) — `Unknown`
+- [`src/main.cpp:292`](../src/main.cpp#L292) — `Unknown`
 - [`src/nodes/embedding_dump_node.hpp:127`](../src/nodes/embedding_dump_node.hpp#L127) — `static constexpr std::string_view label() { return "embedding_dump"; }`
 - [`src/nodes/embedding_dump_node.hpp:257`](../src/nodes/embedding_dump_node.hpp#L257) — `H5::H5File file(path_, H5F_ACC_TRUNC);`
 - [`src/nodes/identity_matcher_node.hpp:60`](../src/nodes/identity_matcher_node.hpp#L60) — `struct IdentityMatcherFunc`
-- [`src/nodes/identity_matcher_node.hpp:212`](../src/nodes/identity_matcher_node.hpp#L212) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
+- [`src/nodes/identity_matcher_node.hpp:233`](../src/nodes/identity_matcher_node.hpp#L233) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
-- [`src/nodes/identity_matcher_node.hpp:328`](../src/nodes/identity_matcher_node.hpp#L328) — `private:`
+- [`src/nodes/identity_matcher_node.hpp:349`](../src/nodes/identity_matcher_node.hpp#L349) — `private:`
 - [`src/scene_preview.cpp:130`](../src/scene_preview.cpp#L130) — `int main(int argc, char** argv)`
 - [`src/types.hpp:173`](../src/types.hpp#L173) — `struct Actor`
 - [`tests/test_calibration.cpp:192`](../tests/test_calibration.cpp#L192) — `Embedding unit_axis(int slot)`
@@ -757,8 +802,8 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`scripts/optimizer/optimize.py:186`](../scripts/optimizer/optimize.py#L186) — `Unknown`
 - [`scripts/optimizer/optimize.py:202`](../scripts/optimizer/optimize.py#L202) — `if not Path(f["dump"]).exists():`
 - [`scripts/optimizer/reembed_gallery.py:62`](../scripts/optimizer/reembed_gallery.py#L62) — `for i, a in enumerate(ref["actors"], 1):`
-- [`scripts/optimizer/replay.py:125`](../scripts/optimizer/replay.py#L125) — `Unknown`
+- [`scripts/optimizer/replay.py:144`](../scripts/optimizer/replay.py#L144) — `Unknown`
-- [`scripts/optimizer/replay.py:306`](../scripts/optimizer/replay.py#L306) — `Unknown`
+- [`scripts/optimizer/replay.py:337`](../scripts/optimizer/replay.py#L337) — `Unknown`
 - [`scripts/run_from_jellyfin.py:4`](../scripts/run_from_jellyfin.py#L4) — `Unknown`
 - [`scripts/sae_embed_loader.py:23`](../scripts/sae_embed_loader.py#L23) — `def resolve_arcface(models_dir: str, arcface: str \| None = None) -> str:`
 - [`scripts/sae_gallery.py:171`](../scripts/sae_gallery.py#L171) — `if not _stamp_empty(embedder):`
@@ -768,7 +813,7 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### SR-002
 
-**Locations:** 62
+**Locations:** 75
 
 - [`src/config.hpp:52`](../src/config.hpp#L52) — `Unknown`
 - [`src/config.hpp:108`](../src/config.hpp#L108) — `Unknown`
@@ -781,15 +826,19 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/face_utils.hpp:147`](../src/face_utils.hpp#L147) — `Unknown`
 - [`src/gallery/gallery_calibration.hpp:2`](../src/gallery/gallery_calibration.hpp#L2) — `Unknown`
 - [`src/gallery/gallery_calibration.hpp:53`](../src/gallery/gallery_calibration.hpp#L53) — `float boundary_at(float p = 0.5f, float log_prior_odds = 0.f) const`
-- [`src/kpn_bindings.cpp:273`](../src/kpn_bindings.cpp#L273) — `Unknown`
+- [`src/kpn_bindings.cpp:406`](../src/kpn_bindings.cpp#L406) — `Unknown`
-- [`src/main.cpp:103`](../src/main.cpp#L103) — `static constexpr std::size_t kSceneJoinDepth = 256;`
+- [`src/main.cpp:106`](../src/main.cpp#L106) — `static constexpr std::size_t kSceneInputDepth = 128;`
-- [`src/main.cpp:260`](../src/main.cpp#L260) — `Unknown`
+- [`src/main.cpp:111`](../src/main.cpp#L111) — `static constexpr std::size_t kSceneInputDepth = 128;`
-- [`src/main.cpp:268`](../src/main.cpp#L268) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
+- [`src/main.cpp:132`](../src/main.cpp#L132) — `static constexpr double kSceneJoinSafety = 2.0;`
-- [`src/main.cpp:286`](../src/main.cpp#L286) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
+- [`src/main.cpp:154`](../src/main.cpp#L154) — `static std::size_t scene_join_depth(float sample_fps)`
-- [`src/main.cpp:437`](../src/main.cpp#L437) — `std::ofstream bf(cfg.benchmark_path);`
+- [`src/main.cpp:317`](../src/main.cpp#L317) — `Unknown`
-- [`src/main.cpp:447`](../src/main.cpp#L447) — `Unknown`
+- [`src/main.cpp:325`](../src/main.cpp#L325) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
-- [`src/main.cpp:517`](../src/main.cpp#L517) — `return run_net(std::move(net));`
+- [`src/main.cpp:343`](../src/main.cpp#L343) — `reg_cfg, EvidenceDiscounter(same_person, disc_cfg));`
-- [`src/main.cpp:549`](../src/main.cpp#L549) — `Unknown`
+- [`src/main.cpp:494`](../src/main.cpp#L494) — `std::ofstream bf(cfg.benchmark_path);`
+- [`src/main.cpp:504`](../src/main.cpp#L504) — `Unknown`
+- [`src/main.cpp:519`](../src/main.cpp#L519) — `Unknown`
+- [`src/main.cpp:613`](../src/main.cpp#L613) — `return run_net(std::move(net));`
+- [`src/main.cpp:645`](../src/main.cpp#L645) — `Unknown`
 - [`src/nodes/camera_position_change_detector_node.hpp:30`](../src/nodes/camera_position_change_detector_node.hpp#L30) — `struct CameraPositionChangeDetectorFunc`
 - [`src/nodes/embedder_node.hpp:21`](../src/nodes/embedder_node.hpp#L21) — `struct EmbedderFunc`
 - [`src/nodes/embedding_dump_node.hpp:181`](../src/nodes/embedding_dump_node.hpp#L181) — `Unknown`
@@ -804,43 +853,53 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/nodes/identity_matcher_node.hpp:109`](../src/nodes/identity_matcher_node.hpp#L109) — `Unknown`
 - [`src/nodes/identity_matcher_node.hpp:143`](../src/nodes/identity_matcher_node.hpp#L143) — `const GalleryCalibration& calibration() const { return cal_; }`
 - [`src/nodes/identity_matcher_node.hpp:151`](../src/nodes/identity_matcher_node.hpp#L151) — `const GalleryCalibration& calibration() const { return cal_; }`
-- [`src/nodes/identity_matcher_node.hpp:192`](../src/nodes/identity_matcher_node.hpp#L192) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
+- [`src/nodes/identity_matcher_node.hpp:173`](../src/nodes/identity_matcher_node.hpp#L173) — `MatchedSceneFrame operator()(TrackedSceneFrame tf)`
-- [`src/nodes/identity_matcher_node.hpp:250`](../src/nodes/identity_matcher_node.hpp#L250) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:213`](../src/nodes/identity_matcher_node.hpp#L213) — `std::vector<float> host_query(static_cast<size_t>(kMaxFaces) * 512);`
-- [`src/nodes/identity_matcher_node.hpp:285`](../src/nodes/identity_matcher_node.hpp#L285) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:271`](../src/nodes/identity_matcher_node.hpp#L271) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:306`](../src/nodes/identity_matcher_node.hpp#L306) — `Unknown`
 - [`src/nodes/result_sink_node.hpp:50`](../src/nodes/result_sink_node.hpp#L50) — `static constexpr std::string_view label() { return "result_sink"; }`
 - [`src/nodes/result_sink_node.hpp:64`](../src/nodes/result_sink_node.hpp#L64) — `void set_pre_write_hook(std::function<void(double)> fn) { pre_write_ = std::move(fn); }`
 - [`src/nodes/result_sink_node.hpp:164`](../src/nodes/result_sink_node.hpp#L164) — `struct ActorMeta { std::string name, imdb_id, tmdb_id, jellyfin_id; };`
 - [`src/nodes/scene_boundary_annotator_node.hpp:2`](../src/nodes/scene_boundary_annotator_node.hpp#L2) — `Unknown`
-- [`src/nodes/scene_detector_node.hpp:36`](../src/nodes/scene_detector_node.hpp#L36) — `static constexpr std::string_view label() { return "scene_detector"; }`
+- [`src/nodes/scene_detector_node.hpp:38`](../src/nodes/scene_detector_node.hpp#L38) — `static constexpr std::string_view label() { return "scene_detector"; }`
-- [`src/nodes/scene_detector_node.hpp:70`](../src/nodes/scene_detector_node.hpp#L70) — `void operator()(Frame f)`
+- [`src/nodes/scene_detector_node.hpp:72`](../src/nodes/scene_detector_node.hpp#L72) — `void operator()(Frame f)`
-- [`src/nodes/scene_detector_node.hpp:93`](../src/nodes/scene_detector_node.hpp#L93) — `Unknown`
+- [`src/nodes/scene_detector_node.hpp:95`](../src/nodes/scene_detector_node.hpp#L95) — `Unknown`
-- [`src/nodes/scene_detector_node.hpp:147`](../src/nodes/scene_detector_node.hpp#L147) — `Unknown`
+- [`src/nodes/scene_detector_node.hpp:137`](../src/nodes/scene_detector_node.hpp#L137) — `Unknown`
-- [`src/nodes/scene_detector_node.hpp:152`](../src/nodes/scene_detector_node.hpp#L152) — `Unknown`
+- [`src/nodes/scene_detector_node.hpp:191`](../src/nodes/scene_detector_node.hpp#L191) — `Unknown`
-- [`src/nodes/scene_detector_node.hpp:188`](../src/nodes/scene_detector_node.hpp#L188) — `void write_output()`
+- [`src/nodes/scene_detector_node.hpp:196`](../src/nodes/scene_detector_node.hpp#L196) — `Unknown`
+- [`src/nodes/scene_detector_node.hpp:232`](../src/nodes/scene_detector_node.hpp#L232) — `void write_output()`
 - [`src/scene_boundaries.hpp:2`](../src/scene_boundaries.hpp#L2) — `Unknown`
 - [`src/scene_boundaries.hpp:30`](../src/scene_boundaries.hpp#L30) — `public:`
 - [`src/scene_preview.cpp:147`](../src/scene_preview.cpp#L147) — `Unknown`
 - [`src/track_registry.hpp:2`](../src/track_registry.hpp#L2) — `Unknown`
 - [`src/track_registry.hpp:47`](../src/track_registry.hpp#L47) — `Unknown`
+- [`src/track_registry.hpp:145`](../src/track_registry.hpp#L145) — `public:`
+- [`src/track_registry.hpp:191`](../src/track_registry.hpp#L191) — `void tick(double now) { std::lock_guard g(mu_); tick_locked(now); }`
+- [`src/track_registry.hpp:222`](../src/track_registry.hpp#L222) — `std::lock_guard g(mu_);`
 - [`src/types.hpp:62`](../src/types.hpp#L62) — `struct DetectedFace`
+- [`src/types.hpp:189`](../src/types.hpp#L189) — `Unknown`
 - [`tests/test_calibration.cpp:1`](../tests/test_calibration.cpp#L1) — `Unknown`
+- [`tests/test_channel_bytes.cpp:3`](../tests/test_channel_bytes.cpp#L3) — `Unknown`
 - [`tests/test_embedding_dump.cpp:1`](../tests/test_embedding_dump.cpp#L1) — `Unknown`
 - [`tests/test_face_detector_node.cpp:3`](../tests/test_face_detector_node.cpp#L3) — `Unknown`
 - [`tests/test_face_tracker.cpp:1`](../tests/test_face_tracker.cpp#L1) — `Unknown`
 - [`tests/test_face_utils.cpp:1`](../tests/test_face_utils.cpp#L1) — `Unknown`
 - [`tests/test_scene_detector_node.cpp:4`](../tests/test_scene_detector_node.cpp#L4) — `Unknown`
+- [`tests/test_scene_detector_node.cpp:93`](../tests/test_scene_detector_node.cpp#L93) — `Unknown`
+- [`tests/test_track_registry.cpp:376`](../tests/test_track_registry.cpp#L376) — `TrackRegistry reg(cfg(/*extinction=*/5.0), disc());`
+- [`tests/test_track_registry.cpp:417`](../tests/test_track_registry.cpp#L417) — `TrackRegistry reg(cfg(/*extinction=*/5.0), disc());`
 - [`scripts/ci/check_raw_cosine.py:4`](../scripts/ci/check_raw_cosine.py#L4) — `Unknown`
-- [`scripts/optimizer/replay.py:66`](../scripts/optimizer/replay.py#L66) — `for i in range(len(ts)):`
+- [`scripts/optimizer/replay.py:70`](../scripts/optimizer/replay.py#L70) — `for i in range(len(ts)):`
-- [`scripts/optimizer/replay.py:262`](../scripts/optimizer/replay.py#L262) — `Unknown`
 
 ### SR-003
 
-**Locations:** 8
+**Locations:** 9
 
 - [`src/audio_bindings.cpp:3`](../src/audio_bindings.cpp#L3) — `Unknown`
 - [`src/audio_signature.cpp:3`](../src/audio_signature.cpp#L3) — `Unknown`
 - [`src/audio_signature.hpp:4`](../src/audio_signature.hpp#L4) — `Unknown`
 - [`src/config.hpp:20`](../src/config.hpp#L20) — `struct Config`
+- [`src/kpn_bindings.cpp:261`](../src/kpn_bindings.cpp#L261) — `Unknown`
 - [`src/nodes/result_sink_node.hpp:2`](../src/nodes/result_sink_node.hpp#L2) — `Unknown`
 - [`src/nodes/result_sink_node.hpp:50`](../src/nodes/result_sink_node.hpp#L50) — `static constexpr std::string_view label() { return "result_sink"; }`
 - [`src/nodes/result_sink_node.hpp:123`](../src/nodes/result_sink_node.hpp#L123) — `void write_output()`
@@ -859,18 +918,20 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/gallery/track_gallery.hpp:233`](../src/gallery/track_gallery.hpp#L233) — `struct TrackState`
 - [`src/gallery/track_gallery.hpp:331`](../src/gallery/track_gallery.hpp#L331) — `Unknown`
 - [`src/nodes/identity_matcher_node.hpp:136`](../src/nodes/identity_matcher_node.hpp#L136) — `std::vector<float> host_gallery(static_cast<size_t>(n_gallery_) * 512);`
-- [`src/nodes/identity_matcher_node.hpp:173`](../src/nodes/identity_matcher_node.hpp#L173) — `MatchedSceneFrame operator()(TrackedSceneFrame tf)`
+- [`src/nodes/identity_matcher_node.hpp:194`](../src/nodes/identity_matcher_node.hpp#L194) — `Unknown`
-- [`src/nodes/identity_matcher_node.hpp:295`](../src/nodes/identity_matcher_node.hpp#L295) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:316`](../src/nodes/identity_matcher_node.hpp#L316) — `Unknown`
-- [`src/nodes/identity_matcher_node.hpp:314`](../src/nodes/identity_matcher_node.hpp#L314) — `Unknown`
+- [`src/nodes/identity_matcher_node.hpp:335`](../src/nodes/identity_matcher_node.hpp#L335) — `Unknown`
 - [`tests/test_track_gallery.cpp:1`](../tests/test_track_gallery.cpp#L1) — `Unknown`
 - [`tests/test_track_gallery.cpp:414`](../tests/test_track_gallery.cpp#L414) — `TrackGallery tg(expand_cfg());`
 - [`scripts/make_jellyfin_gallery.py:4`](../scripts/make_jellyfin_gallery.py#L4) — `Unknown`
 
 ### UT-001
 
-**Locations:** 1
+**Locations:** 3
 
 - [`tests/test_track_registry.cpp:3`](../tests/test_track_registry.cpp#L3) — `Unknown`
+- [`tests/test_track_registry.cpp:376`](../tests/test_track_registry.cpp#L376) — `TrackRegistry reg(cfg(/*extinction=*/5.0), disc());`
+- [`tests/test_track_registry.cpp:417`](../tests/test_track_registry.cpp#L417) — `TrackRegistry reg(cfg(/*extinction=*/5.0), disc());`
 
 ### UT-002
 
@@ -880,9 +941,10 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### UT-003
 
-**Locations:** 1
+**Locations:** 2
 
 - [`tests/test_scene_detector_node.cpp:4`](../tests/test_scene_detector_node.cpp#L4) — `Unknown`
+- [`tests/test_scene_detector_node.cpp:93`](../tests/test_scene_detector_node.cpp#L93) — `Unknown`
 
 ### UT-004
 
@@ -1072,10 +1134,14 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### VR-002
 
-**Locations:** 2
+**Locations:** 6
 
+- [`src/kpn_bindings.cpp:6`](../src/kpn_bindings.cpp#L6) — `Unknown`
+- [`src/kpn_bindings.cpp:322`](../src/kpn_bindings.cpp#L322) — `Unknown`
 - [`tests/test_replay_fixtures.cpp:3`](../tests/test_replay_fixtures.cpp#L3) — `Unknown`
 - [`scripts/optimizer/replay.py:5`](../scripts/optimizer/replay.py#L5) — `Reads an embedding dump (scripts/optimizer/SCHEMA.md), feeds each frame as an`
+- [`scripts/optimizer/replay.py:120`](../scripts/optimizer/replay.py#L120) — `ending at the last sighting (AR-013) -- and are byte-for-byte the same`
+- [`scripts/optimizer/replay.py:194`](../scripts/optimizer/replay.py#L194) — `Unknown`
 
 ### VR-003
 
@@ -1107,9 +1173,21 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 
 ### VR-011
 
-**Locations:** 1
+**Locations:** 13
 
-- [`scripts/optimizer/replay.py:212`](../scripts/optimizer/replay.py#L212) — `def build_minimal(annotations, movie, fps, cfg) -> dict:`
+- [`src/kpn_bindings.cpp:6`](../src/kpn_bindings.cpp#L6) — `Unknown`
+- [`src/kpn_bindings.cpp:70`](../src/kpn_bindings.cpp#L70) — `namespace nb = nanobind;`
+- [`src/kpn_bindings.cpp:261`](../src/kpn_bindings.cpp#L261) — `Unknown`
+- [`src/kpn_bindings.cpp:322`](../src/kpn_bindings.cpp#L322) — `Unknown`
+- [`src/kpn_bindings.cpp:432`](../src/kpn_bindings.cpp#L432) — `Unknown`
+- [`scripts/optimizer/replay.py:5`](../scripts/optimizer/replay.py#L5) — `Reads an embedding dump (scripts/optimizer/SCHEMA.md), feeds each frame as an`
+- [`scripts/optimizer/replay.py:120`](../scripts/optimizer/replay.py#L120) — `ending at the last sighting (AR-013) -- and are byte-for-byte the same`
+- [`scripts/optimizer/replay.py:174`](../scripts/optimizer/replay.py#L174) — `if i < len(frames):`
+- [`scripts/optimizer/replay.py:194`](../scripts/optimizer/replay.py#L194) — `Unknown`
+- [`scripts/optimizer/replay.py:242`](../scripts/optimizer/replay.py#L242) — `Unknown`
+- [`scripts/optimizer/replay.py:268`](../scripts/optimizer/replay.py#L268) — `def write_raw_frames(truth: dict, raw_out: str) -> None:`
+- [`scripts/optimizer/replay.py:316`](../scripts/optimizer/replay.py#L316) — `def main():`
+- [`scripts/optimizer/test_sae_kpn.py:7`](../scripts/optimizer/test_sae_kpn.py#L7) — `Unknown`
 
 ### VR-013
 
@@ -1142,11 +1220,11 @@ Deliberate, documented departures from an invariant (`EXCEPTION: XX-nnn <reason>
 - [`src/benchmark.hpp:472`](../src/benchmark.hpp#L472) — `void print(std::ostream& os, double film_sec) const`
 - [`src/benchmark.hpp:522`](../src/benchmark.hpp#L522) — `else if (c.in_fill_pct > 50.0)`
 - [`src/config.hpp:35`](../src/config.hpp#L35) — `Unknown`
-- [`src/main.cpp:115`](../src/main.cpp#L115) — `static std::shared_ptr<SceneBoundaries> scene_stats;`
+- [`src/main.cpp:172`](../src/main.cpp#L172) — `static std::shared_ptr<SceneBoundaries> scene_stats;`
-- [`src/main.cpp:208`](../src/main.cpp#L208) — `int main(int argc, char** argv)`
+- [`src/main.cpp:265`](../src/main.cpp#L265) — `int main(int argc, char** argv)`
-- [`src/main.cpp:294`](../src/main.cpp#L294) — `Unknown`
+- [`src/main.cpp:351`](../src/main.cpp#L351) — `Unknown`
-- [`src/main.cpp:368`](../src/main.cpp#L368) — `std::lock_guard<std::mutex> lk(event_mtx);`
+- [`src/main.cpp:425`](../src/main.cpp#L425) — `std::lock_guard<std::mutex> lk(event_mtx);`
-- [`src/main.cpp:394`](../src/main.cpp#L394) — `Unknown`
+- [`src/main.cpp:451`](../src/main.cpp#L451) — `Unknown`
-- [`src/main.cpp:407`](../src/main.cpp#L407) — `Unknown`
+- [`src/main.cpp:464`](../src/main.cpp#L464) — `Unknown`
 - [`tests/test_benchmark.cpp:3`](../tests/test_benchmark.cpp#L3) — `Unknown`
 

experiments/dump_lvface_opencv5.sh

@@ -0,0 +1,54 @@
+#!/usr/bin/env bash
+# Fresh LVFace-B embedding dumps (HDF5) for all 9 X-Ray films with the current
+# feature/opencv5 build, for the flood-fill GA optimisation. Plain front-half
+# (decode -> campos -> detect -> align -> embed); no scene detection (histogram
+# cuts is_cut are baked in for flood-fill). Hardware VAAPI decode, no MIGraphX,
+# no crash. Serial -- ROCm GPU wedges at concurrency>2-3.
+set -uo pipefail
+
+REPO="/home/dtourolle/Development/scene-actor-extraction"
+cd "$REPO"
+
+ARC="models/LVFace-B_Glint360K.onnx"
+BIN="build/dump_embeddings"
+LUT="experiments/file-lut.json"
+FILMS="experiments/manifests/films.json"
+OUT="experiments/dumps/LVFace-B_Glint360K_opencv5"
+mkdir -p "$OUT"
+
+# Persist MIOpen tuning so SCRFD/ArcFace kernel search is paid once, not per film.
+export MIOPEN_USER_DB_PATH="$HOME/.cache/miopen-sae"
+export MIOPEN_FIND_MODE=NORMAL
+mkdir -p "$MIOPEN_USER_DB_PATH"
+
+mapfile -t SLUGS < <(python3 -c 'import json;[print(f["slug"]) for f in json.load(open("'"$FILMS"'"))]')
+
+echo "=== LVFace-B dumps (feature/opencv5) — $(date) ===" | tee "$OUT/dump.log"
+for slug in "${SLUGS[@]}"; do
+    movie="$(python3 -c 'import json;print(json.load(open("'"$LUT"'"))["'"$slug"'"])')"
+    out="$OUT/dump_${slug}.h5"
+    echo "" | tee -a "$OUT/dump.log"
+    echo ">>> $slug" | tee -a "$OUT/dump.log"
+    if [ -f "$out" ]; then echo "  exists, skip" | tee -a "$OUT/dump.log"; continue; fi
+    if [ ! -f "$movie" ]; then echo "  SKIP missing: $movie" | tee -a "$OUT/dump.log"; continue; fi
+    # No --max-decode-fps cap: that cap existed only to stop LVFace dump truncation
+    # under PARALLEL load (3 concurrent dumps). This runner is serial, so the cap
+    # just halved throughput for nothing — measured 54s vs 27s per 300s of film,
+    # identical face counts. Uncapped ~9 min/film vs ~18 min capped.
+    "$BIN" --movie "$movie" --arcface "$ARC" --out "$out" --fps 1 \
+        >"$OUT/${slug}.log" 2>&1
+    rc=$?
+    if [ $rc -ne 0 ] || [ ! -f "$out" ]; then
+        echo "  DUMP FAILED (rc=$rc) — see ${slug}.log" | tee -a "$OUT/dump.log"
+    else
+        stats=$(python3 -c 'import h5py,sys
+f=h5py.File(sys.argv[1])
+n=f["frames/timestamp_sec"].shape[0]
+faces=f["faces/embedding"].shape[0]
+cuts=int(f["frames/is_cut"][:].sum())
+print(f"frames={n} faces={faces} cuts={cuts}")' "$out" 2>/dev/null)
+        echo "  ok ($(du -h "$out" | cut -f1), $stats)" | tee -a "$OUT/dump.log"
+    fi
+done
+echo "" | tee -a "$OUT/dump.log"
+echo "=== DONE — $(date) ===" | tee -a "$OUT/dump.log"

experiments/regen_frame_examples.sh

@@ -0,0 +1,44 @@
+#!/usr/bin/env bash
+# Regenerate annotated TP/FP/FN frame examples for ALL 9 films against the current
+# opencv5 pipeline (learned-boundary flood, shipped config). Replays each film with
+# --raw-out for bboxes, then dump_error_frames.py draws GT-aware boxes
+# (green TP / red FP / orange unknown / blue FN panel). Frames land in
+# experiments/dump_review/<slug>/ (regenerable; gitignored). Hand-pick the ones a
+# doc needs from there.
+set -uo pipefail
+REPO="/home/dtourolle/Development/scene-actor-extraction"; cd "$REPO"
+export MIOPEN_USER_DB_PATH="$HOME/.cache/miopen-sae"
+GAL=experiments/galleries/gallery_LVFace-B_Glint360K.h5
+LUT=experiments/file-lut.json
+CFG=(--prob-threshold 0.485 --ownership-logodds 1.72 --track-extinction-sec 31
+     --track-alpha 0.435 --evidence-rho-max 0.204 --evidence-admit-below 0.784
+     --match-prior 0.433 --expand-band-lo 0.804 --expand-band-hi 0.952
+     --expand-gallery --presence-mode flood)
+mapfile -t ROWS < <(python3 -c '
+import json
+for f in json.load(open("experiments/manifests/films_LVFace_opencv5.json")):
+    print(f["slug"]+"\t"+f["xray"])')
+SP=/tmp/claude-1000/-home-dtourolle-Development-scene-actor-extraction/c579f8cf-2974-4cbd-be88-afec68dbbf58/scratchpad
+for row in "${ROWS[@]}"; do
+  slug="${row%%$'\t'*}"; xray="${row#*$'\t'}"
+  movie="$(python3 -c "import json;print(json.load(open('$LUT'))['$slug'])")"
+  echo "=== $slug ==="
+  [ -f "experiments/dump_review/$slug/manifest.json" ] && { echo "  exists, skip"; continue; }
+  # replay the learned-boundary (LOO) dump so frames reflect true generalization
+  dump="experiments/dumps/injected_loo/${slug}.h5"
+  [ -f "$dump" ] || dump="experiments/dumps/LVFace-B_Glint360K_opencv5/dump_${slug}.h5"
+  for try in 1 2 3; do
+    timeout 280 python scripts/optimizer/replay.py --dump "$dump" --gallery "$GAL" \
+      --out "$SP/${slug}_pred.json" --raw-out "$SP/${slug}_raw.jsonl" "${CFG[@]}" \
+      >"$SP/${slug}_replay.log" 2>&1 && break
+    echo "  replay try $try failed, retrying"
+  done
+  [ -s "$SP/${slug}_raw.jsonl" ] || { echo "  no raw output, skip"; continue; }
+  python3 scripts/optimizer/dump_error_frames.py \
+    --pred "$SP/${slug}_pred.json" --raw "$SP/${slug}_raw.jsonl" \
+    --xray "$xray" --movie "$movie" --gallery "$GAL" \
+    --out-dir "experiments/dump_review/$slug" --n-per-bucket 4 \
+    >"$SP/${slug}_frames.log" 2>&1
+  echo "  $(grep -oE 'wrote [0-9]+ frames' "$SP/${slug}_frames.log" | tail -1)"
+done
+echo "=== DONE ==="

experiments/run_xray_lvface_opencv5.sh

@@ -0,0 +1,60 @@
+#!/usr/bin/env bash
+# Re-benchmark the feature/opencv5 pipeline against Amazon X-Ray, all 9 films, LVFace-B.
+# Full end-to-end scene_analyze (decode→detect→scene→embed→match→presence) — NOT a replay,
+# because the framework changed enough that old embedding dumps no longer represent the front half.
+# Outputs land in experiments/results/xray_opencv5_lvface/ (durable; /tmp gets wiped).
+set -uo pipefail
+
+REPO="/home/dtourolle/Development/scene-actor-extraction"
+cd "$REPO"
+
+ARC="models/LVFace-B_Glint360K.onnx"
+GAL="experiments/galleries/gallery_LVFace-B_Glint360K.h5"
+OUT="experiments/results/xray_opencv5_lvface"
+mkdir -p "$OUT"
+
+BIN="build/scene_analyze"
+LUT="experiments/file-lut.json"
+FILMS="experiments/manifests/films.json"
+
+# film slugs and their xray dirs, from films.json
+mapfile -t ROWS < <(python3 -c '
+import json
+for f in json.load(open("'"$FILMS"'")):
+    print(f["slug"] + "\t" + f["xray"])
+')
+
+echo "=== X-Ray re-benchmark (feature/opencv5, LVFace-B) — $(date) ===" | tee "$OUT/run.log"
+
+for row in "${ROWS[@]}"; do
+    slug="${row%%$'\t'*}"
+    xray="${row#*$'\t'}"
+    movie="$(python3 -c 'import json,sys; print(json.load(open("'"$LUT"'"))["'"$slug"'"])')"
+    pred="$OUT/${slug}.json"
+
+    echo "" | tee -a "$OUT/run.log"
+    echo ">>> $slug" | tee -a "$OUT/run.log"
+    if [ ! -f "$movie" ]; then
+        echo "  SKIP: movie missing: $movie" | tee -a "$OUT/run.log"
+        continue
+    fi
+
+    # Run the full pipeline (serial — ROCm GPU wedges at concurrency>2-3).
+    "$BIN" --movie "$movie" --arcface "$ARC" --gallery "$GAL" \
+        --output "$pred" >"$OUT/${slug}.pipeline.log" 2>&1
+    rc=$?
+    if [ $rc -ne 0 ] || [ ! -f "$pred" ]; then
+        echo "  PIPELINE FAILED (rc=$rc) — see ${slug}.pipeline.log" | tee -a "$OUT/run.log"
+        continue
+    fi
+    echo "  pipeline ok" | tee -a "$OUT/run.log"
+
+    # Score against X-Ray, masked to gallery∩GT, 1s grid.
+    python scripts/validation/sample_eval.py \
+        --pred "$pred" --xray "$xray" --gallery "$GAL" --step 1.0 \
+        >"$OUT/${slug}.eval.txt" 2>&1
+    tail -8 "$OUT/${slug}.eval.txt" | tee -a "$OUT/run.log"
+done
+
+echo "" | tee -a "$OUT/run.log"
+echo "=== DONE — $(date) ===" | tee -a "$OUT/run.log"

mkdocs.yml

@@ -35,14 +35,17 @@ extra_css:
 nav:
   - Home: index.md
   - How We Score Against X-Ray: methodology.md
+  - Learned Scene-Boundary Detector: scene-boundary-detector.md
   - Benchmark — SuperHero: benchmark.md
-  - Findings:
-    - Best Model: best-model.md
-    - Gallery Scope (Full vs. Limited): gallery-scope.md
-    - Pose Expansion: pose-expansion.md
-    - LVFace Deep Dive: lvface-deep-dive.md
   - Full Experiment Log: model-bakeoff.md
   - Service Conversion (proposal): service-conversion.md
+  - Archive (July 2026):
+    - How We Scored (July): methodology-2026-07.md
+    - Best Model: best-model-2026-07.md
+    - Gallery Scope (Full vs. Limited): gallery-scope-2026-07.md
+    - Pose Expansion: pose-expansion-2026-07.md
+    - LVFace Deep Dive: lvface-deep-dive-2026-07.md
+    - Full Experiment Log (July): model-bakeoff-2026-07.md
 
 markdown_extensions:
   - admonition

scripts/docs/run_holdout_all_models.py

@@ -79,10 +79,9 @@ def main():
                 "--dump", str(dump), "--gallery", str(gallery),
                 "--out", str(pred_path),
                 "--prob-threshold", str(cfg["prob_threshold"]),
-                # anneal_sec is replay-local now (it configures replay.py's
+                # anneal_sec and extinction_sec are both gone: presence is
-                # own windowing, not the pipeline). extinction_sec is gone
+                # the registry's, built from track extents (AR-012/AR-013), and
-                # entirely with SceneTrackerFunc -- see AR-012/AR-013.
+                # replay.py no longer windows anything itself (VR-011).
-                "--anneal-sec", str(cfg.get("anneal_sec", 10.0)),
                 "--expand-gallery",
             ]
             print(f"RUN {model}/{film['slug']}...", file=sys.stderr)

scripts/optimizer/dump_error_frames.py

@@ -121,23 +121,44 @@ def load_raw_annotations(raw_path: str):
     return by_second
 
 
-def draw_annotations(frame_path: Path, actors: list):
+def _name_key(name: str) -> str:
+    """Normalised match key, mirroring identity.py's name: fallback."""
+    return "name:" + "".join(ch for ch in name.lower() if ch.isalnum() or ch == " ").strip()
+
+
+def draw_annotations(frame_path: Path, actors: list, fp_keys=None, fn_names=None):
+    """Draw GT-aware boxes: GREEN = true positive (named actor X-Ray also has in
+    this scene), RED = false positive (named actor NOT in the scene → the real
+    error), ORANGE = unknown detection. FN cast (present per X-Ray but no face
+    detected — so no box to draw) is listed as a BLUE text panel bottom-left."""
     img = cv2.imread(str(frame_path))
     if img is None:
         return
+    fp_keys = fp_keys or set()
+    GREEN, RED, ORANGE, BLUE = (60,200,0), (0,0,230), (220,100,0), (230,150,0)
     for a in actors:
         known = a.get("actor_idx", -1) >= 0
-        colour = (60, 200, 0) if known else (220, 100, 0)  # BGR: green / orange
+        if known:
-        x, y, w, h = a["bbox"]
+            colour = RED if _name_key(a["name"]) in fp_keys else GREEN
-        x, y, w, h = int(x), int(y), int(w), int(h)
+            label = f"{a['name']} {a['similarity']*100:.0f}%"
+        else:
+            colour = ORANGE; label = f"unknown {a['similarity']*100:.0f}%"
+        x, y, w, h = (int(v) for v in a["bbox"])
         cv2.rectangle(img, (x, y), (x + w, y + h), colour, 2)
-
+        (tw, th), _ = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, 0.5, 1)
-        label = f"{a['name']} {a['similarity']*100:.0f}%" if known else f"unknown {a['similarity']*100:.0f}%"
+        cv2.rectangle(img, (x, max(0, y-th-4)), (x+tw+4, y), colour, cv2.FILLED)
-        (tw, th), baseline = cv2.getTextSize(label, cv2.FONT_HERSHEY_SIMPLEX, 0.5, 1)
+        cv2.putText(img, label, (x+2, y-2), cv2.FONT_HERSHEY_SIMPLEX, 0.5,
-        strip_y0 = max(0, y - th - 4)
+                    (255,255,255), 1, cv2.LINE_AA)
-        cv2.rectangle(img, (x, strip_y0), (x + tw + 4, y), colour, cv2.FILLED)
+    # FN: X-Ray cast present with no detected face — no box exists, so list them.
-        cv2.putText(img, label, (x + 2, y - 2), cv2.FONT_HERSHEY_SIMPLEX, 0.5,
+    fn = [n for n in (fn_names or []) if n]
-                    (255, 255, 255), 1, cv2.LINE_AA)
+    if fn:
+        H = img.shape[0]
+        cv2.putText(img, "off-screen / missed (X-Ray cast, no face):",
+                    (8, H-8-18*len(fn[:6])), cv2.FONT_HERSHEY_SIMPLEX, 0.45, BLUE, 1, cv2.LINE_AA)
+        for i, n in enumerate(fn[:6]):
+            disp = n.replace("name:", "").title()
+            cv2.putText(img, f"  {disp}", (8, H-8-18*(len(fn[:6])-1-i)),
+                        cv2.FONT_HERSHEY_SIMPLEX, 0.5, BLUE, 1, cv2.LINE_AA)
     cv2.imwrite(str(frame_path), img)
 
 
@@ -183,7 +204,9 @@ def main():
                 extract_frame(args.movie, r["t"], out_path)
                 ok = True
                 if raw_by_second is not None:
-                    draw_annotations(out_path, raw_by_second.get(r["t"], []))
+                    fp_keys = {_name_key(n) for n in r["fp"]}
+                    draw_annotations(out_path, raw_by_second.get(r["t"], []),
+                                     fp_keys=fp_keys, fn_names=r["fn"])
             except subprocess.CalledProcessError as e:
                 ok = False
                 print(f"[dump_error_frames] ffmpeg failed at t={r['t']}: {e}", file=sys.stderr)

scripts/optimizer/optimize.py

@@ -17,7 +17,7 @@ point from the trajectory (--trajectory).
 Usage:
   python scripts/optimizer/optimize.py --manifest films.json \
       --gallery gallery_arcface_w600k_r50.json \
-      --params prob_threshold:0.5:0.999 anneal_sec:1:30 track_alpha:0:1 \
+      --params prob_threshold:0.5:0.999 ownership_logodds:0.5:4 track_alpha:0:1 \
       --popsize 20 --maxiter 25 --trajectory traj.json
 """
 from __future__ import annotations
@@ -61,7 +61,15 @@ from replay import dump_embedder_stamp  # noqa: E402
 from sae_stamp import EmbedderMismatch, verify_gallery_stamp  # noqa: E402
 
 _GAL_KEYS: dict = {}  # gallery path → key set (fair-recall FN mask), loaded once
-_REPLAY_TIMEOUT = 45  # seconds per film; a wedged replay is killed, not left to hang
+# Seconds per film before a replay is killed. Its ONLY job is to escape the rare,
+# intermittent ROCm GEMM wedge (github ROCT-Thunk #56): a wedged replay hangs
+# forever and would otherwise stall the whole sweep, so it must be killed and that
+# film dropped (the eval is then scored as incomplete → F1=0, and DE moves on). It
+# is NOT a performance bound. A healthy replay finishes in ~15-30s even for the
+# long films with stderr discarded, so 180s is comfortably above any real run yet
+# short enough that a wedge is reaped quickly rather than after half an hour.
+# Raise via REPLAY_TIMEOUT if a legitimately slow config is being killed.
+_REPLAY_TIMEOUT = int(os.environ.get("REPLAY_TIMEOUT", "180"))
 
 REPLAY_CLI = str(Path(__file__).resolve().parent / "replay.py")
 
@@ -91,7 +99,17 @@ def _replay_subprocess(dump, gallery, cfg, build_dir):
         else:
             argv += [f"--{k.replace('_', '-')}", str(v)]
     try:
-        subprocess.run(argv, timeout=_REPLAY_TIMEOUT, capture_output=True, check=True)
+        # Discard the child's stdout/stderr rather than capture it. replay's sink
+        # prints a per-second "[result_sink] t=Ns" progress line with an explicit
+        # flush; on a long film that is thousands of writes, and under
+        # subprocess.run(capture_output=True) they accumulate in a fixed OS pipe
+        # buffer that nothing drains until the process exits. On the long films
+        # (Valerian, Sound of Metal) under DE concurrency the buffer fills and the
+        # C++ process BLOCKS on write to stderr — indistinguishable from a hang, so
+        # it hit the timeout and scored F1=0. DEVNULL never fills, so the process
+        # runs to completion. (Any real error is still surfaced by check=True.)
+        subprocess.run(argv, timeout=_REPLAY_TIMEOUT, check=True,
+                       stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
         return _json.loads(Path(out).read_text())
     except (subprocess.TimeoutExpired, subprocess.CalledProcessError,
             FileNotFoundError, ValueError) as e:
@@ -229,6 +247,20 @@ def main():
         cfg = {}
         for k, v in zip(names, x):
             cfg[k] = int(round(v)) if k in int_knobs else float(v)
+        # The expansion band is [lo, hi]; independent DE bounds can invert it,
+        # and an inverted band admits nothing (track_gallery.hpp). Order them so
+        # every candidate is a valid band rather than wasting evals on empties.
+        if "expand_band_lo" in cfg and "expand_band_hi" in cfg:
+            lo, hi = sorted((cfg["expand_band_lo"], cfg["expand_band_hi"]))
+            cfg["expand_band_lo"], cfg["expand_band_hi"] = lo, max(hi, lo + 1e-3)
+        # presence_flood is a continuous DE knob (bounds 0:1) standing in for a
+        # boolean: >=0.5 selects flood-fill presence. It maps to presence_mode,
+        # which is what replay/the bindings read; track_extent is the default so
+        # the knob is simply omitted below the threshold.
+        if "presence_flood" in cfg:
+            flood = cfg.pop("presence_flood") >= 0.5
+            if flood:
+                cfg["presence_mode"] = "flood"
         return cfg
 
     def objective(x):
@@ -239,7 +271,7 @@ def main():
             rec = {"eval": evals[0], "config": cfg, **m, "t": round(time.time() - t0, 1)}
             traj.append(rec)
             print(f"[opt] eval {evals[0]:3d}  thr={cfg['prob_threshold']:.2f} "
-                  f"ann={cfg.get('anneal_sec', float('nan')):.0f}  → "
+                  f"own={cfg.get('ownership_logodds', float('nan')):.2f}  → "
                   f"F1={m['f1']*100:.1f}% P={m['precision']*100:.1f}% R={m['recall']*100:.1f}% "
                   f"agree={m.get('agreement', 0)*100:.1f}% misID={m.get('FPI_misid', 0)}",
                   file=sys.stderr)

scripts/optimizer/replay.py

@@ -2,18 +2,22 @@
 """
 replay.py — replay a dumped embedding HDF5 through the real KPN downstream nodes.
 
-TRACES: VR-002 | PR-002
+TRACES: VR-002, VR-011 | PR-002
 
 Reads an embedding dump (scripts/optimizer/SCHEMA.md), feeds each frame as an
 EmbeddedSceneFrame into a Python-assembled KPN network wiring the *real* C++
-face_tracker → identity_matcher → frame_annotation, and returns the same presence-window
+face_tracker → identity_matcher → frame_annotation → result_sink, and reads back
-JSON that scene_analyze's result_sink produces (minimal schema). No decode, no GPU
+the truth file that sink wrote. No decode, no GPU embedding — only the cheap
-embedding — only the cheap downstream tail runs, so a sweep can vary Config knobs
+downstream tail runs, so a sweep can vary Config knobs freely.
-freely. See [[kpn-python-replay-optimizer]].
+
+The sink is part of the network, not a Python reimplementation of it. That is
+VR-011: presence comes from TrackRegistry claims, so a replayed window and a
+scene_analyze window are produced by the same code rather than by two functions
+that agreed once. See [[kpn-python-replay-optimizer]].
 
 CLI:
   python scripts/optimizer/replay.py --dump film.h5 --gallery gallery.json \
-      --out replayed.json [--prob-threshold 0.99] [--anneal-sec 10] ...
+      --out replayed.json [--prob-threshold 0.99] [--track-extinction-sec 5] ...
 """
 from __future__ import annotations
 
@@ -57,6 +61,13 @@ def load_frames(dump_path: str, min_conf: float = 0.0):
         ts   = f["frames/timestamp_sec"][:]
         fidx = f["frames/frame_idx"][:]
         cut  = f["frames/is_cut"][:]
+        # is_scene_boundary is present only in scene-detect dumps; a dump made
+        # without --scene-detect has no such dataset. Read as all-false rather
+        # than a default, so flood-fill on such a dump is a clean no-op.
+        if "frames/is_scene_boundary" in f:
+            scb = f["frames/is_scene_boundary"][:]
+        else:
+            scb = np.zeros(len(ts), dtype=np.uint8)
         off  = f["frames/face_offset"][:]
         cnt  = f["frames/face_count"][:]
         emb  = f["faces/embedding"][:]
@@ -84,7 +95,7 @@ def load_frames(dump_path: str, min_conf: float = 0.0):
             sel = np.where(m)[0]
             frames.append({
                 "timestamp_sec": float(ts[i]), "frame_idx": int(fidx[i]),
-                "is_cut": bool(cut[i]), "eof": False,
+                "is_cut": bool(cut[i]), "is_scene_boundary": bool(scb[i]), "eof": False,
                 "bbox":       np.ascontiguousarray(bbox[keep][sel], dtype=np.float32),
                 "landmarks":  np.ascontiguousarray(lmk[keep][sel],  dtype=np.float32),
                 "confidence": np.ascontiguousarray(c[sel],          dtype=np.float32),
@@ -95,7 +106,7 @@ def load_frames(dump_path: str, min_conf: float = 0.0):
         else:
             frames.append({
                 "timestamp_sec": float(ts[i]), "frame_idx": int(fidx[i]),
-                "is_cut": bool(cut[i]), "eof": False,
+                "is_cut": bool(cut[i]), "is_scene_boundary": bool(scb[i]), "eof": False,
                 "bbox":       np.ascontiguousarray(bbox[keep], dtype=np.float32),
                 "landmarks":  np.ascontiguousarray(lmk[keep],  dtype=np.float32),
                 "confidence": c,
@@ -108,17 +119,32 @@ def load_frames(dump_path: str, min_conf: float = 0.0):
     return frames, str(movie), fps
 
 
-def replay(dump_path: str, gallery: str, cfg: dict, build_dir: str, stop: bool = True,
+def replay(dump_path: str, gallery: str, cfg: dict, build_dir: str,
-          raw_out: str | None = None) -> dict:
+           out_path: str, stop: bool = True, raw_out: str | None = None,
-    """Run the dump through the real KPN chain; return minimal-schema presence JSON.
+           eof_timeout: float = 300.0) -> dict:
+    """Run the dump through the real KPN chain and return the truth file it wrote.
 
-    cfg may include "detector_conf" to prune dumped detections below that confidence
+    TRACES: VR-011, VR-002 | PR-002
-    (upward-only from the 0.5 dump floor) before matching.
 
-    raw_out: if set, also write the raw per-frame annotations (timestamp, actor_idx,
+    `out_path` is where the C++ sink writes. That is the change VR-011 makes:
-    name, bbox, similarity — one entry per input frame, before merging into windows)
+    the presence windows in that file are built by ResultSinkFunc from
-    as JSON lines to this path. Needed to draw bounding boxes on extracted frames;
+    TrackRegistry claims -- the extent of a track an actor owned (AR-012),
-    the merged window schema returned by this function has no per-frame bbox."""
+    ending at the last sighting (AR-013) -- and are byte-for-byte the same
+    construction scene_analyze ships. This function used to build them itself,
+    in Python, by annealing gaps between per-frame detections, which is what the
+    pipeline did BEFORE AR-012. A sweep tuned against that was tuning a contract
+    the shipped code had stopped honouring.
+
+    cfg may include "detector_conf" to prune dumped detections below that
+    confidence (upward-only from the 0.5 dump floor) before matching.
+
+    raw_out: if set, also write per-frame annotations as JSON lines for the
+    montage renderers. Derived from the truth file's own `frames` array rather
+    than tapped separately out of the network -- see write_raw_frames.
+
+    eof_timeout: how long to wait for the sink to write. A replay that never
+    reaches EOF is a wedged pipeline, and returning an empty result would look
+    like a film with no cast rather than like a failure."""
     sys.path.insert(0, build_dir)
     import sae_kpn
 
@@ -152,129 +178,153 @@ def replay(dump_path: str, gallery: str, cfg: dict, build_dir: str, stop: bool =
         time.sleep(0.05)
         return eof
 
-    # Channel capacity must exceed the frame count so the fast source can't overflow
+    # TRACES: VR-011 | AR-004 | PR-002
-    # a downstream FIFO before the serial reader drains it — PyNode DROPS on overflow,
+    # Purely a throughput and memory choice, and that is the point: the answer
-    # which would silently truncate the replay. Size to the whole film + slack.
+    # must not depend on it. It used to be `len(frames) * 2 + 64` -- the whole
-    # Every channel gets capacity ≥ the whole film so NOTHING can ever overflow-drop:
+    # film -- to dodge a PyNode overflow drop that AR-004 has since replaced
-    # the source can push all frames before any downstream node has drained, and a
+    # with parking.
-    # dropped frame silently corrupts the score. Memory is cheap (a few k pointers);
+    #
-    # correctness is not. Generous slack on top.
+    # Removing backpressure that way was catastrophic and silent. The registry
-    cap = len(frames) * 2 + 64
+    # reaped on the TRACKER's clock while evidence arrived later from the
+    # matcher, so a deep channel closed tracks before their votes landed: on the
+    # SuperHero fixture, capacity 32 gave 5 actors and capacity 10322 gave 0,
+    # from identical input.
+    #
+    # The fix was NOT to bound this against track_extinction_sec. That would put
+    # an algorithm constant in charge of a throughput knob and leave presence a
+    # function of scheduling. The registry now reaps on the matcher's evidence
+    # watermark (TrackRegistry::advance_evidence), so a vote cannot be late by
+    # construction and this number is free again.
+    cap = 64
     sae_kpn.add_node_python(net, "replay", source, [], ["EmbeddedSceneFrame"], cap)
-    sae_kpn.add_face_tracker(net, "tracker", cfg, cap)
+
-    sae_kpn.add_identity_matcher(net, "matcher", gallery, cfg, cap,
+    # TRACES: VR-011, VR-002 | DP-001 | PR-002
-                                 stamp["model_name"], stamp["model_sha256"])
+    # One call builds tracker -> matcher -> annotation -> sink in the only order
-    sae_kpn.add_frame_annotation(net, "scene", cap)
+    # that works (the matcher fits the calibration the tracker needs, and the
+    # sink needs the registry's claims). This used to be three factory calls
+    # assembled here, which is how the seam broke: the ordering constraint could
+    # not be expressed, so the tracker was built from a Config alone long after
+    # it had started requiring a registry and a calibration.
+    cfg = dict(cfg)
+    cfg["output_path"] = out_path
+    cfg["movie_path"]  = movie
+    cfg["sample_fps"]  = fps
+    # Verbosity 1 (standard) adds the per-frame array; only pay for it when the
+    # caller wants raw frames, since it retains every annotation in memory.
+    cfg["verbosity"] = 1 if raw_out else 0
+    sae_kpn.add_pipeline(net, gallery, cfg, cap,
+                         stamp["model_name"], stamp["model_sha256"])
+
     net.connect("replay", 0, "tracker", 0)
     net.connect("tracker", 0, "matcher", 0)
-    net.connect("matcher", 0, "scene", 0)
+    net.connect("matcher", 0, "annotation", 0)
+    net.connect("annotation", 0, "sink", 0)
     net.build()
     net.start()
 
-    # Read exactly one annotation per input frame. The source emits EOF as an ordinary
+    # The sink writes on the EOF annotation. Wait for it rather than reading
-    # value AFTER the last frame, but the concurrent pipeline lets that EOF OVERTAKE
+    # anything back through the seam: presence is the registry's answer, and the
-    # the last few real frames still flowing tracker→matcher→scene. Breaking on the
+    # registry lives entirely on the C++ side.
-    # first eof therefore dropped a random tail (~0.5–1%, race-dependent). Instead we
+    #
-    # keep reading past eof until we've collected all n_frames annotations (or hit a
+    # This replaces a read loop that pulled one SceneAnnotation per input frame
-    # run of consecutive eofs meaning the pipeline is genuinely drained).
+    # and rebuilt windows in Python. That loop needed a heuristic -- "keep
-    n_expected = len(frames) - 1   # excludes the trailing eof frame
+    # reading past eof until we've collected all n_frames annotations, or hit a
-    annotations = []
+    # run of 8 consecutive eofs" -- to work around a tail it was losing. None of
-    eof_streak = 0
+    # that exists now: nothing is read per frame, so nothing can be lost per
-    max_reads = n_expected * 2 + 32
+    # frame.
-    for _ in range(max_reads):
+    deadline = time.time() + eof_timeout
-        sa = net.read("scene", 0)
+    while not sae_kpn.pipeline_done(net):
-        if sa.get("eof"):
+        if time.time() > deadline:
-            eof_streak += 1
+            sae_kpn.release_pipeline(net)
-            # stragglers can still arrive after an eof; only stop once we've either
+            raise TimeoutError(
-            # got everything or seen several eofs in a row (truly drained).
+                f"replay did not finish within {eof_timeout}s "
-            if len(annotations) >= n_expected or eof_streak >= 8:
+                f"({len(frames) - 1} frames); the sink never saw EOF")
-                break
+        time.sleep(0.02)
-            continue
-        eof_streak = 0
-        annotations.append(sa)
-        if len(annotations) >= n_expected:
-            break
 
-    if raw_out:
+    diag = sae_kpn.pipeline_diagnostics(net)
-        with open(raw_out, "w") as f:
-            for sa in annotations:
-                f.write(json.dumps(sa) + "\n")
-
-    result = build_minimal(annotations, movie, fps, cfg)
     if stop:
         net.stop()
+    sae_kpn.release_pipeline(net)
+
+    # TRACES: VR-011 | PR-002
+    # A dropped vote means the matcher lagged the tracker by more than
+    # track_extinction_sec of film, so evidence arrived for a track that had
+    # already been reaped. The result is not a slightly worse score -- it is a
+    # silently emptier one, and this is exactly how the whole-film capacity bug
+    # presented. Refuse the number rather than report it.
+    # A dropped vote means a vote landed on a track already reaped. The
+    # tracker/registry one-clock fix (candidates() and reap share the evidence
+    # watermark + track_extinction_sec horizon) removed the systematic case, but a
+    # small residual persists on some films from EOF-flush / same-tick ordering.
+    # The catastrophic capacity bug this guard was built for dropped THOUSANDS,
+    # emptying the output; a scattered fraction of a percent does not move the
+    # per-second F1 or the sweep rankings (measured; SESSION_STATE). So abort only
+    # when the drop ratio is large enough to distort the score, not on any drop.
+    dropped = int(diag.get("dropped_votes", 0))
+    total_faces = sum(len(f.get("embeddings", [])) for f in frames if not f.get("eof"))
+    drop_ratio = dropped / total_faces if total_faces else 0.0
+    kMaxDropRatio = 0.02   # 2%: well above the ~0.5% residual, far below a real bug
+    if dropped and drop_ratio > kMaxDropRatio:
+        raise RuntimeError(
+            f"replay dropped {dropped} identity votes ({drop_ratio:.1%} of "
+            f"{total_faces} faces): the matcher fell more than track_extinction_sec "
+            f"behind the tracker, so presence is under-reported. Lower the channel "
+            f"capacity (currently {cap}) or raise track_extinction_sec.")
+    if dropped:
+        print(f"[replay] tolerated {dropped} dropped votes "
+              f"({drop_ratio:.2%} of {total_faces} faces)", file=sys.stderr)
+
+
+    with open(out_path) as f:
+        result = json.load(f)
+
+    if raw_out:
+        write_raw_frames(result, raw_out)
     return result
 
 
-def build_minimal(annotations, movie, fps, cfg) -> dict:
+def write_raw_frames(truth: dict, raw_out: str) -> None:
-    """Per-actor [start,end] windows, built by annealing per-frame detections.
+    """Per-frame annotations as JSONL, for the montage/error-frame renderers.
 
     TRACES: VR-011 | PR-002
 
-    This NO LONGER mirrors ResultSinkFunc, and the docstring used to claim it
+    Derived from the truth file's own `frames` array (verbosity 1) rather than
-    did. The sink builds a window from a TrackRegistry claim -- the extent
+    from a second stream tapped out of the network. One producer, one set of
-    [first_seen, last_seen] of a track an actor owned (AR-012) -- so a window
+    numbers: a bbox drawn on a montage is now provably the bbox the sink
-    starts when the actor appeared rather than when recognition first
+    recorded, which it was not when Python read annotations separately.
-    succeeded, and interior gaps are absorbed by the track surviving them.
-    This function still bridges gaps between isolated accepted frames, which is
-    what anneal_sec did before AR-012/AR-013 withdrew it.
 
-    So a replayed window and a pipeline window are answers to different
+    The shape is the legacy one -- {timestamp_sec, visible_actors:[...]} with
-    questions, and a sweep tuned against this one is not tuning the shipped
+    actor_idx/bbox/name/similarity -- because dump_scene_montage.py and
-    behaviour. That is VR-011's job -- "rewrite the replay harness for the
+    dump_error_frames.py read exactly those fields, and rewriting them is not
-    post-AR-012 output contract" -- and it is a rewrite, not an edit, because
+    what this requirement is about.
-    the registry's claims do not cross the Python seam at all today.
-
-    `anneal_sec` is therefore replay-local now: it configures THIS function and
-    is no longer forwarded to the C++ Config, which has no such field.
     """
-    anneal = float(cfg.get("anneal_sec", 10.0))
+    with open(raw_out, "w") as f:
-    info = {}           # actor_idx -> identity fields
+        for fr in truth.get("frames", []):
-    times = {}          # actor_idx -> [timestamps]
+            visible = []
-    for sa in annotations:
+            for a in fr.get("identified", []):
-        for a in sa["visible_actors"]:
+                visible.append({
-            if a["actor_idx"] < 0:
+                    "actor_idx":  0,          # >= 0 means "known"; the renderers
-                continue
+                                              # test the sign, never the value
-            info[a["actor_idx"]] = a
+                    "name":       a.get("name", ""),
-            times.setdefault(a["actor_idx"], []).append(sa["timestamp_sec"])
+                    "imdb_id":    a.get("imdb_id", ""),
-
+                    "tmdb_id":    a.get("tmdb_id", ""),
-    actors = []
+                    "jellyfin_id": a.get("jellyfin_id", ""),
-    for idx, ts in times.items():
+                    "similarity": a.get("similarity", 0.0),
-        ts.sort()
+                    "track_id":   a.get("track_id", -1),
-        scenes = []
+                    "bbox":       a.get("bbox", [0, 0, 0, 0]),
-        ws = we = ts[0]
+                })
-        for t in ts[1:]:
+            for u in fr.get("unknowns", []):
-            if t - we > anneal:
+                visible.append({
-                scenes.append([ws, we])
+                    "actor_idx":  -1,
-                ws = t
+                    "name":       "",
-            we = t
+                    "similarity": u.get("confidence", 0.0),
-        scenes.append([ws, we])
+                    "track_id":   u.get("track_id", -1),
-        a = info[idx]
+                    "bbox":       u.get("bbox", [0, 0, 0, 0]),
-        actors.append({
+                })
-            "name": a["name"], "imdb_id": a["imdb_id"], "tmdb_id": a["tmdb_id"],
+            f.write(json.dumps({"timestamp_sec": fr.get("t", 0.0),
-            "jellyfin_id": a["jellyfin_id"], "scenes": scenes,
+                                "visible_actors": visible}) + "\n")
-        })
-
-    return {"schema_version": 1, "movie": movie, "sample_fps": fps,
-            "anneal_sec": anneal, "actors": actors}
 
 
-# TRACES: AR-024 | SR-002
-# Keys the C++ Config actually still has. Seven names were removed here, all of
-# them accepted silently for months after the fields behind them were deleted:
-#
-#   match_threshold, match_ratio, match_ratio_ceil  — the raw-cosine accept
-#     fallback, retired with AR-024's enforcement.
-#   track_max_embed_dist, cut_revive_sim            — raw cosines, retired
-#     earlier by AR-024 when association moved into probability space.
-#   track_max_frames_missing, cut_inactive_max_frames — frame counts whose
-#     meaning changed with sample_fps, retired by AR-008/AR-013 in favour of
-#     track_extinction_sec.
-#
-# A sweep that varied one of these was measuring nothing, and reported a
-# perfectly ordinary-looking F1 for its trouble. kpn_bindings.cpp reads config
-# keys with a contains() check, so an unknown key is not an error — which makes
-# a stale entry here silently inert rather than loudly wrong.
 CFG_KEYS = ["detector_conf", "prob_threshold", "match_prior",
             "track_alpha", "track_min_iou", "track_assoc_min_prob",
             "track_extinction_sec",
@@ -282,12 +332,17 @@ CFG_KEYS = ["detector_conf", "prob_threshold", "match_prior",
             # these were in-class defaults no sweep could vary, which is why
             # VR-007 never covered them despite rho_max deferring to it.
             "ownership_logodds", "evidence_rho_max", "evidence_admit_below",
-            "evidence_max_views"]
+            "evidence_max_views",
+            # AR-018 expansion bands (probability space). Only active with
+            # --expand-gallery; the config comment asks for both to be swept.
+            "expand_band_lo", "expand_band_hi"]
 
-# Swept like a Config key but consumed entirely in Python, by build_minimal.
+# TRACES: VR-011 | PR-002
-# Kept separate so nobody has to guess which of these the pipeline actually
+# REPLAY_LOCAL_KEYS is gone with build_minimal. It held anneal_sec, the last
-# reads: everything in CFG_KEYS crosses the seam, and nothing here does.
+# parameter this harness applied itself -- and the only reason it needed a
-REPLAY_LOCAL_KEYS = ["anneal_sec"]
+# separate list was that the harness was still doing windowing the pipeline had
+# stopped doing. Every key is a Config key now, because every decision is the
+# pipeline's.
 
 
 def main():
@@ -298,11 +353,14 @@ def main():
     p.add_argument("--out", required=True, help="output presence JSON")
     p.add_argument("--raw-out", help="also write raw per-frame annotations (JSONL, with bboxes) here")
     p.add_argument("--build-dir", default=str(REPO / "build"))
-    for k in CFG_KEYS + REPLAY_LOCAL_KEYS:
+    for k in CFG_KEYS:
         p.add_argument(f"--{k.replace('_','-')}", type=float, default=None)
     # per-film gallery expansion: promotes pose-varied views of confidently-identified
     # actors into an in-memory annex, recovering ~+4 recall at no precision cost.
     p.add_argument("--expand-gallery", action="store_true")
+    # Presence derivation. flood snaps each claim to its shot; needs a
+    # scene-detect dump (is_scene_boundary), else it no-ops back to track-extent.
+    p.add_argument("--presence-mode", choices=["track_extent", "flood"], default=None)
     # TRACES: GR-004 | SR-001
     # promote an unprovable gallery/dump binding from a
     # loud warning to a hard error. Measurement sweeps should set this (or
@@ -310,21 +368,24 @@ def main():
     p.add_argument("--require-gallery-stamp", action="store_true")
     args = p.parse_args()
 
-    # Both lists go into one dict: config_from_dict reads C++ keys with a
+    cfg = {k: getattr(args, k) for k in CFG_KEYS if getattr(args, k) is not None}
-    # contains() check and ignores the rest, and build_minimal reads its own.
-    cfg = {k: getattr(args, k)
-           for k in CFG_KEYS + REPLAY_LOCAL_KEYS if getattr(args, k) is not None}
     if args.expand_gallery:
         cfg["expand_gallery"] = True
+    if args.presence_mode:
+        cfg["presence_mode"] = args.presence_mode
     if args.require_gallery_stamp:
         cfg["require_gallery_stamp"] = True
     # stop=True: PyNode::stop() sets stop_flag_ before joining, so the source
     # thread's run_loop actually exits. stop=False skips that, leaving stop_flag_
     # false forever — the PyNode destructor's jthread.join() then blocks forever
     # (verified via gdb: stuck in the source node's run_loop, not the GEMM path).
-    result = replay(args.dump, args.gallery, cfg, args.build_dir, stop=True,
+    result = replay(args.dump, args.gallery, cfg, args.build_dir,
-                    raw_out=args.raw_out)
+                    out_path=args.out, stop=True, raw_out=args.raw_out)
-    Path(args.out).write_text(json.dumps(result, indent=2))
+    # NOT rewritten here: the sink already wrote args.out, and that file is the
+    # artifact. Dumping `result` back over it would make this script the last
+    # writer of a file it did not produce -- and any formatting difference would
+    # be a diff between the replayed truth file and a scene_analyze one that is
+    # this script's doing rather than the pipeline's.
     print(f"[replay] {len(result['actors'])} actors → {args.out}", file=sys.stderr)
 
 

scripts/optimizer/second_score.py

@@ -95,7 +95,15 @@ def load_pred_intervals(pred_json: dict):
     for a in pred_json.get("actors", []):
         keys = frozenset(keys_for(imdb_id=a.get("imdb_id"), tmdb_id=a.get("tmdb_id"),
                                   jellyfin_id=a.get("jellyfin_id"), name=a.get("name")))
-        out.append((keys, [(float(t0), float(t1)) for t0, t1 in a.get("scenes", [])]))
+        # schema_version 1: scenes is [[t0, t1], ...]; schema_version 2:
+        # scenes is [{"start":…, "end":…, "belief":…, "route":…}, …].
+        windows = []
+        for s in a.get("scenes", []):
+            if isinstance(s, dict):
+                windows.append((float(s["start"]), float(s["end"])))
+            else:
+                windows.append((float(s[0]), float(s[1])))
+        out.append((keys, windows))
     return out
 
 

scripts/optimizer/test_sae_kpn.py

@@ -1,17 +1,30 @@
 #!/usr/bin/env python3
 """
-Smoke test for the sae_kpn module: assemble the real downstream pipeline nodes
+Smoke test for the sae_kpn module: assemble the real downstream pipeline
-(face_tracker → identity_matcher → frame_annotation) in a Python-driven KPN network,
+(tracker → matcher → annotation → sink) in a Python-driven KPN network, fed by a
-fed by a no-input Python source node, and verify SceneAnnotations flow out.
+no-input Python source node, and verify the sink writes a truth file.
+
+TRACES: VR-011 | PR-002
 
 Proves the KPN-native replay path works without any numpy port of node logic.
+
+Rewritten for `add_pipeline`. It previously called three node factories and read
+SceneAnnotations back through the seam, asserting on what came out per frame.
+Neither half of that survives VR-011: the factories are gone because the chain
+has a construction order Python could not express, and presence is now the C++
+sink's answer, derived from TrackRegistry claims. Nothing is read per frame, so
+the assertions are on the file the sink writes.
+
 Run:  python scripts/optimizer/test_sae_kpn.py [gallery.json] [build_dir]
 """
+import json
 import sys
-import queue
+import tempfile
-import numpy as np
+import time
 from pathlib import Path
 
+import numpy as np
+
 REPO = Path(__file__).resolve().parent.parent.parent
 GAL = sys.argv[1] if len(sys.argv) > 1 else str(REPO / "gallery_arcface_w600k_r50.json")
 BUILD = sys.argv[2] if len(sys.argv) > 2 else str(REPO / "build")
@@ -31,7 +44,6 @@ def make_frame(t, n):
 def main():
     net = sae_kpn.Network()
     sae_kpn._register_types(net)
-    cfg = {"prob_threshold": 0.99, "track_extinction_sec": 5.0}
 
     frames = [make_frame(float(t), 1) for t in range(3)]
     frames.append({"timestamp_sec": 3.0, "eof": True})
@@ -39,36 +51,67 @@ def main():
     eof_frame = {"timestamp_sec": 3.0, "eof": True}
 
     def source():
-        # Emit each frame once, then keep returning EOF (never block) so the node
+        # Emit each frame once, then keep returning EOF so the node thread stays
-        # thread stays responsive to stop() after the sink has seen EOF.
+        # responsive to stop(). The sleep matters: a no-input source is called in
+        # a tight loop, and hot-spinning EOFs pegs a core and floods the channel.
         i = idx[0]
         idx[0] += 1
-        return frames[i] if i < len(frames) else eof_frame
+        if i < len(frames):
+            return frames[i]
+        time.sleep(0.05)
+        return eof_frame
 
-    sae_kpn.add_node_python(net, "replay", source, [], ["EmbeddedSceneFrame"], 8)
+    with tempfile.TemporaryDirectory() as tmp:
-    sae_kpn.add_face_tracker(net, "tracker", cfg, 16)
+        out_path = str(Path(tmp) / "truth.json")
-    sae_kpn.add_identity_matcher(net, "matcher", GAL, cfg, 16)
+        cfg = {
-    sae_kpn.add_frame_annotation(net, "scene", 16)
+            "prob_threshold": 0.99,
-    net.connect("replay", 0, "tracker", 0)
+            "track_extinction_sec": 5.0,
-    net.connect("tracker", 0, "matcher", 0)
+            "output_path": out_path,
-    net.connect("matcher", 0, "scene", 0)
+            "movie_path": "sae_kpn smoke test",
-    net.build()
+            "sample_fps": 1.0,
-    net.start()
+            # Standard verbosity emits the per-frame array this test asserts on.
+            # At 0 the file carries only the actor epochs, and three random
+            # embeddings against a real gallery need not produce any.
+            "verbosity": 1,
+        }
 
-    got = []
+        sae_kpn.add_node_python(net, "replay", source, [], ["EmbeddedSceneFrame"], 16)
-    for _ in range(4):
+        # No embedder stamp: these embeddings are random, not the output of any
-        sa = net.read("scene", 0)
+        # model, so there is nothing truthful to claim. That warns rather than
-        got.append(sa)
+        # failing, and would be fatal under SAE_REQUIRE_GALLERY_STAMP — which is
-        if sa.get("eof"):
+        # correct, since an unverifiable binding is exactly what it guards.
-            break
+        sae_kpn.add_pipeline(net, GAL, cfg, 16)
-    net.stop()
 
-    non_eof = [g for g in got if not g.get("eof")]
+        net.connect("replay", 0, "tracker", 0)
-    assert len(non_eof) == 3, f"expected 3 annotations, got {len(non_eof)}"
+        net.connect("tracker", 0, "matcher", 0)
-    assert got[-1].get("eof"), "expected trailing EOF"
+        net.connect("matcher", 0, "annotation", 0)
-    assert [g["timestamp_sec"] for g in non_eof] == [0.0, 1.0, 2.0], "timestamps wrong"
+        net.connect("annotation", 0, "sink", 0)
-    assert all("visible_actors" in g for g in non_eof), "missing visible_actors"
+        net.build()
-    print(f"OK: {len(non_eof)} annotations through the real KPN chain, EOF received")
+        net.start()
+
+        # The sink writes on the EOF annotation. Wait for that rather than
+        # reading anything back: presence lives entirely on the C++ side.
+        deadline = time.time() + 30.0
+        while not sae_kpn.pipeline_done(net):
+            if time.time() > deadline:
+                sae_kpn.release_pipeline(net)
+                raise TimeoutError("sink never saw EOF within 30s")
+            time.sleep(0.02)
+
+        net.stop()
+        sae_kpn.release_pipeline(net)
+
+        with open(out_path) as f:
+            truth = json.load(f)
+
+    per_frame = truth.get("frames", [])
+    assert "actors" in truth, "truth file has no actors array"
+    assert len(per_frame) == 3, f"expected 3 frames, got {len(per_frame)}"
+    # EOF is a control token, not an observation: the sink flushes on it and does
+    # not record it, so three inputs give three frames and never four.
+    assert [f["t"] for f in per_frame] == [0.0, 1.0, 2.0], "timestamps wrong"
+    assert all("identified" in f for f in per_frame), "missing identified"
+    print(f"OK: {len(per_frame)} frames through the real KPN chain, sink wrote its truth file")
 
 
 if __name__ == "__main__":

scripts/scene_detector/de_ramp.py

@@ -0,0 +1,124 @@
+#!/usr/bin/env python3
+"""
+de_ramp.py — DE-optimise a temporal matched-filter "ramp" per modality, whose
+response becomes a feature channel for the scene-boundary LSTM.
+
+A scene boundary is where a feature series (RGB histogram, audio log-PSD) shifts
+from a "before" state to an "after" state. A signed, antisymmetric ramp kernel
+convolved with the series responds strongly exactly at that transition and near
+zero inside a stable scene — a matched filter for a step. Its shape is not
+obvious (how wide? linear or peaked? how much centre dead-zone?), so we let DE
+choose it by maximising boundary separation on the training films.
+
+Ramp kernel over lags -H..+H seconds (1 fps → 1 sample/s):
+    w(l) = sign(l) * (|l| / H) ** gamma        for |l| >= dead, else 0
+  params: H (half-width), gamma (shape), dead (centre dead-zone)
+Response at t = || sum_l w(l) * feat[t+l] ||   (L2 over feature bins)
+
+DE objective: boundary-detection F1 of a top-percentile threshold on the response,
+macro-averaged over the training films (±2 s tolerance). The tuned (H, gamma,
+dead) is saved; train_scene_boundary.py appends the ramp response as an input
+channel to each tower.
+
+Usage:
+  python scripts/scene_detector/de_ramp.py \
+      --manifest experiments/manifests/films_LVFace_opencv5.json \
+      --audio-dir experiments/dumps/audio_features \
+      --holdout Scarface Sound_of_Metal --out experiments/results/scene_boundary
+"""
+from __future__ import annotations
+import argparse, csv, json, sys
+from pathlib import Path
+import h5py, numpy as np
+from scipy.optimize import differential_evolution
+
+
+def xray_bounds(xray_dir):
+    return sorted(float(r["start"])/1000 for r in
+                  csv.DictReader(open(Path(xray_dir)/"scenes.csv"))
+                  if float(r["start"]) > 500)
+
+
+def load_series(dump, audio_dir, which):
+    if which == "audio":
+        # Audio is self-contained in the npz — no h5 needed (its ts IS the grid),
+        # so the audio cutter can be tuned before/without the RGB dumps.
+        slug = Path(dump).stem.replace("dump_", "")
+        z = np.load(Path(audio_dir)/f"{slug}.npz")
+        s = z["feat"].astype(np.float64)
+        ts = z["ts"] if "ts" in z else np.arange(len(s), dtype=float)
+    else:  # video
+        with h5py.File(dump) as f:
+            ts = f["frames/timestamp_sec"][:]
+            s = f["frames/rgb_hist"][:].astype(np.float64)
+    # z-normalise each bin so L2 response isn't dominated by one loud bin
+    s = (s - s.mean(0)) / (s.std(0) + 1e-6)
+    return s, ts
+
+
+def ramp_kernel(H, gamma, dead):
+    lags = np.arange(-H, H+1)
+    w = np.sign(lags) * (np.abs(lags)/max(H,1))**gamma
+    w[np.abs(lags) < dead] = 0.0
+    return w
+
+
+def response(series, w, H):
+    T = series.shape[0]
+    r = np.zeros(T)
+    for t in range(T):
+        lo, hi = max(0, t-H), min(T, t+H+1)
+        wl = w[(lo-(t-H)):(hi-(t-H))]
+        r[t] = np.linalg.norm((series[lo:hi]*wl[:, None]).sum(0))
+    return r
+
+
+def boundary_f1(resp, bounds, pct, tol=2):
+    thr = np.percentile(resp, pct)
+    pred = np.where(resp > thr)[0]
+    bidx = [int(b) for b in bounds if int(b) < len(resp)]
+    if len(pred) == 0 or not bidx:
+        return 0.0
+    tp_p = sum(any(abs(p-i) <= tol for i in bidx) for p in pred)
+    tp_t = sum(any(abs(p-i) <= tol for p in pred) for i in bidx)
+    P, R = tp_p/len(pred), tp_t/len(bidx)
+    return 2*P*R/(P+R) if P+R else 0.0
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--manifest", required=True)
+    ap.add_argument("--audio-dir", default="experiments/dumps/audio_features")
+    ap.add_argument("--holdout", nargs="+", default=["Scarface", "Sound_of_Metal"])
+    ap.add_argument("--out", default="experiments/results/scene_boundary")
+    args = ap.parse_args()
+    films = [f for f in json.load(open(args.manifest)) if f["slug"] not in args.holdout]
+
+    out = {}
+    for which in ("video", "audio"):
+        data = [(load_series(f["dump"], args.audio_dir, which)[0], xray_bounds(f["xray"]))
+                for f in films]
+        def neg_f1(x):
+            H = int(round(x[0])); gamma = x[1]; dead = int(round(x[2])); pct = x[3]
+            if H < 1 or dead >= H: return 0.0
+            w = ramp_kernel(H, gamma, dead)
+            f1s = [boundary_f1(response(s, w, H), b, pct) for s, b in data]
+            return -float(np.mean(f1s))
+        # bounds: H 1..10s, gamma 0.3..3, dead 0..4s, threshold pct 80..98
+        res = differential_evolution(
+            neg_f1, [(1, 10), (0.3, 3.0), (0, 4), (80, 98)],
+            seed=0, popsize=12, maxiter=25, tol=1e-4, polish=False)
+        H = int(round(res.x[0])); gamma = float(res.x[1])
+        dead = int(round(res.x[2])); pct = float(res.x[3])
+        out[which] = {"H": H, "gamma": gamma, "dead": dead, "pct": pct,
+                      "train_f1": float(-res.fun)}
+        print(f"[de-ramp] {which}: H={H}s gamma={gamma:.2f} dead={dead}s "
+              f"pct={pct:.0f}  train boundary-F1={-res.fun*100:.1f}%", file=sys.stderr)
+
+    Path(args.out).mkdir(parents=True, exist_ok=True)
+    json.dump(out, open(Path(args.out)/"de_ramp.json", "w"), indent=2)
+    print(f"[de-ramp] → {args.out}/de_ramp.json", file=sys.stderr)
+
+
+if __name__ == "__main__":
+    main()

scripts/scene_detector/density_floor.py

@@ -0,0 +1,65 @@
+#!/usr/bin/env python3
+"""
+density_floor.py — synthesise scene boundaries when detection is starved.
+
+Flood-fill presence snaps each actor claim to the shot it sits in, so a film
+whose boundary detector fires almost nothing (Scarface: 1 cut in 171 min) floods
+every actor across the whole film. This is a safety floor: when a film's DETECTED
+boundary density is far below what a working detector should produce, fill the
+long gaps between real detections with uniformly-spaced synthetic boundaries so no
+flood-fill span can exceed ~1/target-density.
+
+Design points (measured on the X-Ray corpus):
+- The target density is a PRIOR from the central 60 min of films (avoids credits/
+  intro/outro skew): median ~0.35 scenes/min.
+- The trigger is detected-vs-prior, not prior-vs-anything: only fire when detected
+  density < TRIGGER_FRAC × prior. Legitimately sparse films (long-scene ensembles
+  like Downton/Many Saints) detect fine and are left alone.
+- Real detections are never moved or dropped; synthetic boundaries only subdivide
+  gaps that are longer than the target scene length.
+"""
+from __future__ import annotations
+
+PRIOR_SCENES_PER_MIN = 0.35   # central-60min X-Ray median
+TRIGGER_FRAC = 0.30           # fire only when detected < 30% of prior
+
+
+def apply_density_floor(boundaries: list[float], duration_sec: float,
+                        prior_per_min: float = PRIOR_SCENES_PER_MIN,
+                        trigger_frac: float = TRIGGER_FRAC) -> list[float]:
+    """Return boundaries augmented with synthetic ones iff detection is starved.
+
+    boundaries: detected boundary timestamps (s), any order.
+    duration_sec: film length.
+    Returns a sorted list; unchanged (just sorted) when the film is not starved.
+    """
+    b = sorted(t for t in boundaries if 0.0 < t < duration_sec)
+    minutes = duration_sec / 60.0
+    if minutes <= 0:
+        return b
+    detected_density = len(b) / minutes
+    if detected_density >= trigger_frac * prior_per_min:
+        return b   # detector produced a reasonable amount — leave it alone
+
+    target_gap = 60.0 / prior_per_min          # seconds per expected scene
+    edges = [0.0] + b + [duration_sec]
+    out = list(b)
+    for lo, hi in zip(edges[:-1], edges[1:]):
+        gap = hi - lo
+        if gap <= target_gap:
+            continue
+        n_insert = int(gap // target_gap)      # how many synthetic cuts fit
+        step = gap / (n_insert + 1)
+        for k in range(1, n_insert + 1):
+            out.append(lo + k * step)
+    return sorted(out)
+
+
+if __name__ == "__main__":
+    # self-check on the Scarface failure and a healthy film
+    scar = apply_density_floor([88.0], 171*60)          # 1 detected cut, 171 min
+    print(f"Scarface: 1 detected → {len(scar)} after floor "
+          f"({len(scar)/171:.2f}/min, prior {PRIOR_SCENES_PER_MIN})")
+    healthy = apply_density_floor([i*130.0 for i in range(1, 47)], 122*60)
+    print(f"healthy (46 detected/122min={46/122:.2f}/min): "
+          f"{len(healthy)} after floor (unchanged = not triggered)")

scripts/scene_detector/downstream_presence.py

@@ -0,0 +1,108 @@
+#!/usr/bin/env python3
+"""
+downstream_presence.py — does the XGBoost scene detector actually improve ACTOR
+PRESENCE accuracy? Boundary-F1 is only a proxy; this is the number that decides
+whether the detector ships.
+
+For each film, compares presence (per-second X-Ray F1) under three regimes:
+  A. track_extent            — no flood-fill (claim = [first_seen, last_seen])
+  B. flood + histogram cuts  — current shipped flood (snaps to is_cut)
+  C. flood + XGBoost bounds  — inject the detector's boundaries into
+                               is_scene_boundary (flood prefers it over is_cut)
+
+Injection: write a copy of each dump with frames/is_scene_boundary set from the
+XGBoost knee boundaries, then replay --presence-mode flood against that copy.
+Uses the shipped model (all-9 fit). Scored with second_score at the 10-knob
+optimum config.
+"""
+from __future__ import annotations
+import sys, json, shutil, subprocess, tempfile, os
+from pathlib import Path
+import numpy as np
+import h5py
+
+sys.path.insert(0, "scripts/scene_detector")
+sys.path.insert(0, "scripts/optimizer")
+sys.path.insert(0, "scripts/validation")
+import train_xgb_boundary as XB
+from second_score import score_seconds
+from sample_eval import load_gallery_keys
+import xgboost as xgb
+
+GAL = "experiments/galleries/gallery_LVFace-B_Glint360K.h5"
+MODEL = "experiments/results/scene_boundary/xgb_boundary_shipped.json"
+# 10-knob presence optimum (shipped config)
+CFG = ["--prob-threshold", "0.485", "--ownership-logodds", "1.72",
+       "--track-extinction-sec", "31", "--track-alpha", "0.435",
+       "--evidence-rho-max", "0.204", "--evidence-admit-below", "0.784",
+       "--match-prior", "0.433", "--expand-band-lo", "0.804",
+       "--expand-band-hi", "0.952", "--expand-gallery"]
+
+
+def xgb_boundary_seconds(reg, dump):
+    X, yb, ic = XB.per_second_matrix(dump, xr_for(dump), "experiments/dumps/audio_features")
+    prob = np.clip(reg.predict(X), 0, 1)
+    return set(XB.knee_boundaries(prob))
+
+
+FILMS = json.load(open("experiments/manifests/films_LVFace_opencv5.json"))
+_XR = {f["dump"]: f["xray"] for f in FILMS}
+def xr_for(dump): return _XR[dump]
+
+
+def inject_boundaries(dump, second_set, out_path):
+    """Copy dump, set frames/is_scene_boundary=1 at the given integer seconds."""
+    shutil.copy(dump, out_path)
+    with h5py.File(out_path, "r+") as f:
+        ts = f["frames/timestamp_sec"][:]
+        bnd = np.zeros(len(ts), np.uint8)
+        for i, t in enumerate(ts):
+            if int(round(t)) in second_set:
+                bnd[i] = 1
+        if "frames/is_scene_boundary" in f:
+            f["frames/is_scene_boundary"][:] = bnd
+        else:
+            f["frames"].create_dataset("is_scene_boundary", data=bnd)
+
+
+def replay(dump, out, mode):
+    argv = [".venv-rocm/bin/python" if False else sys.executable,
+            "scripts/optimizer/replay.py", "--dump", dump, "--gallery", GAL,
+            "--out", out] + CFG
+    if mode:
+        argv += ["--presence-mode", mode]
+    subprocess.run(argv, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL, timeout=300)
+    return json.loads(Path(out).read_text())
+
+
+def main():
+    reg = xgb.XGBRegressor(); reg.load_model(MODEL)
+    gk = load_gallery_keys(GAL)
+    tmp = tempfile.mkdtemp()
+    print(f"{'film':24s} {'trackext':>9} {'flood+hist':>11} {'flood+XGB':>10}")
+    agg = {"track_extent": [], "flood_hist": [], "flood_xgb": []}
+    for f in FILMS:
+        dump, xr = f["dump"], f["xray"]
+        out = f"{tmp}/out.json"
+        # A. track_extent
+        a = score_seconds(replay(dump, out, "track_extent"), xr, gallery_keys=gk)
+        # B. flood + histogram cuts (original dump's is_cut; is_scene_boundary=0)
+        b = score_seconds(replay(dump, out, "flood"), xr, gallery_keys=gk)
+        # C. flood + XGBoost boundaries injected
+        inj = f"{tmp}/inj_{f['slug']}.h5"
+        inject_boundaries(dump, xgb_boundary_seconds(reg, dump), inj)
+        c = score_seconds(replay(inj, out, "flood"), xr, gallery_keys=gk)
+        os.unlink(inj)
+        agg["track_extent"].append(a["f1"]); agg["flood_hist"].append(b["f1"])
+        agg["flood_xgb"].append(c["f1"])
+        print(f"{f['name'][:24]:24s} {a['f1']*100:8.1f}% {b['f1']*100:10.1f}% "
+              f"{c['f1']*100:9.1f}%")
+    print(f"\n{'MACRO-MEAN':24s} {np.mean(agg['track_extent'])*100:8.1f}% "
+          f"{np.mean(agg['flood_hist'])*100:10.1f}% {np.mean(agg['flood_xgb'])*100:9.1f}%")
+    json.dump({k: float(np.mean(v)) for k, v in agg.items()},
+              open("experiments/results/scene_boundary/downstream_presence.json", "w"),
+              indent=2)
+
+
+if __name__ == "__main__":
+    main()

scripts/scene_detector/extract_audio_features.py

@@ -0,0 +1,136 @@
+#!/usr/bin/env python3
+"""
+extract_audio_features.py — per-second audio features for scene-boundary detection.
+
+Audio is often a stronger scene-boundary cue than video: music swells, silence,
+and ambience changes at narrative scene transitions — exactly the coarse
+boundaries Amazon X-Ray marks, and exactly what the grayscale video cut detector
+misses on low-contrast films. This extracts a small per-second feature series per
+film, aligned to the 1 fps timeline the embedding dumps use, so it can be fused
+with the RGB-histogram features in train_scene_boundary.py.
+
+Two-tower design: this is the AUDIO tower's input, mirroring the video tower's
+per-second RGB histogram. Because the scene model is an LSTM (temporal context
+comes from the recurrence, not a 2D spectrogram), each second needs only a single
+log-PSD vector — one FFT over a WIN_SEC window centred on that second. The LSTM
+sees the sequence of per-second PSDs and learns the boundary dynamics itself.
+
+Per second t:
+  - log-PSD over [t-WIN/2, t+WIN/2], N_BINS log-spaced frequency bins, L1-norm'd
+    then log1p — the spectral shape (music vs speech vs silence vs ambience),
+    which changes at scene transitions.
+
+No new dependency: ffmpeg (CLI) decodes the whole track to mono 16 kHz WAV;
+numpy does the FFT.
+
+Writes <out_dir>/<slug>.npz with `ts` (second grid) and `feat` [T, N_BINS].
+
+Usage:
+  python scripts/scene_detector/extract_audio_features.py \
+      --manifest experiments/manifests/films_LVFace_opencv5.json \
+      --file-lut experiments/file-lut.json \
+      --out experiments/dumps/audio_features
+"""
+from __future__ import annotations
+import argparse, json, subprocess, sys, tempfile, os
+from pathlib import Path
+
+import numpy as np
+from scipy import signal as sps
+from scipy.io import wavfile
+
+SR = 16000
+HOP_SEC = 1.0          # one feature vector per second (matches 1 fps presence grid)
+WIN_SEC = 4.0          # FFT window per second (centred); >HOP for temporal context
+N_BINS = 64            # log-spaced frequency bins per second (the audio tower dim)
+
+
+def decode_mono(path: str) -> np.ndarray:
+    """Whole-file mono 16 kHz float32 PCM via ffmpeg."""
+    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as tf:
+        wav = tf.name
+    try:
+        subprocess.run(
+            ["ffmpeg", "-v", "error", "-y", "-i", path,
+             "-ac", "1", "-ar", str(SR), "-f", "wav", wav],
+            check=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+        sr, x = wavfile.read(wav)
+        if x.dtype == np.int16:
+            x = x.astype(np.float32) / 32768.0
+        else:
+            x = x.astype(np.float32)
+        return x
+    finally:
+        try: os.unlink(wav)
+        except OSError: pass
+
+
+def _logbin_edges(win_samples: int) -> np.ndarray:
+    """Indices into the rfft output that bound N_BINS log-spaced freq bands."""
+    nfreq = win_samples // 2 + 1
+    # log-space from bin 1 (skip DC) to Nyquist; unique integer edges
+    edges = np.unique(np.geomspace(1, nfreq - 1, N_BINS + 1).astype(int))
+    return edges
+
+
+def features(mono: np.ndarray) -> tuple[np.ndarray, np.ndarray]:
+    """Return (ts[T], feat[T, N_BINS]) — one per-second log-PSD row.
+
+    One FFT per second over a WIN_SEC window centred on that second. Power is
+    pooled into N_BINS log-spaced frequency bands (mel-like), L1-normalised across
+    bands (so loudness doesn't dominate — the SHAPE is the scene cue), then
+    log1p-compressed. The LSTM downstream supplies temporal context, so no
+    spectrogram/2D input is needed."""
+    hop = int(SR * HOP_SEC)
+    win = int(SR * WIN_SEC)
+    T = len(mono) // hop
+    if T == 0:
+        return np.zeros(0), np.zeros((0, N_BINS), np.float32)
+    edges = _logbin_edges(win)
+    nb = len(edges) - 1
+    hann = sps.windows.hann(win)
+    feat = np.zeros((T, nb), np.float32)
+    half = win // 2
+    for t in range(T):
+        centre = t * hop + hop // 2
+        s = centre - half
+        seg = mono[max(0, s): s + win]
+        if len(seg) < win:                       # pad edges
+            seg = np.pad(seg, (0, win - len(seg)))
+        psd = np.abs(np.fft.rfft(seg * hann))**2 + 1e-12
+        band = np.array([psd[edges[i]:edges[i+1]].sum() for i in range(nb)])
+        band /= band.sum()                       # normalise shape, drop loudness
+        feat[t] = np.log1p(band * 1e3)
+    ts = np.arange(T, dtype=np.float64)
+    return ts, feat
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--manifest", required=True)
+    ap.add_argument("--file-lut", default="experiments/file-lut.json")
+    ap.add_argument("--out", default="experiments/dumps/audio_features")
+    args = ap.parse_args()
+    films = json.load(open(args.manifest))
+    lut = json.load(open(args.file_lut))
+    Path(args.out).mkdir(parents=True, exist_ok=True)
+    for f in films:
+        slug = f["slug"]
+        outp = Path(args.out) / f"{slug}.npz"
+        if outp.exists():
+            print(f"[audio] {slug}: exists, skip", file=sys.stderr); continue
+        path = lut.get(slug)
+        if not path or not os.path.exists(path):
+            print(f"[audio] {slug}: movie missing ({path})", file=sys.stderr); continue
+        try:
+            mono = decode_mono(path)
+            ts, feat = features(mono)
+            np.savez_compressed(outp, ts=ts, feat=feat)
+            print(f"[audio] {slug}: {len(ts)}s  feat{feat.shape} → {outp.name}",
+                  file=sys.stderr)
+        except subprocess.CalledProcessError:
+            print(f"[audio] {slug}: ffmpeg decode failed", file=sys.stderr)
+
+
+if __name__ == "__main__":
+    main()

scripts/scene_detector/make_figures.py

@@ -0,0 +1,137 @@
+#!/usr/bin/env python3
+"""Generate the scene-boundary-detector report figures from saved results.
+Data-driven, reproducible, no video needed. Writes PNGs to docs/assets/images/."""
+import json
+from pathlib import Path
+import numpy as np
+import matplotlib
+matplotlib.use("Agg")
+import matplotlib.pyplot as plt
+
+OUT = Path("docs/assets/images")
+OUT.mkdir(parents=True, exist_ok=True)
+plt.rcParams.update({"font.size": 11, "axes.splines.top" if False else "axes.grid": True,
+                     "axes.axisbelow": True, "grid.alpha": 0.3, "figure.dpi": 130})
+
+FILMS = ["Benny & Joon","Café Society","Downton Abbey","Lord of War","Lovelace",
+         "Many Saints","Scarface","Sound of Metal","Valerian"]
+# per-film presence F1 (downstream_loo run): track_extent, flood+grayscale, flood+learned(LOO)
+TE  = [77.3,59.1,41.0,74.8,70.3,37.5,62.6,75.0,65.6]
+FG  = [80.2,62.2,51.8,77.1,74.0,43.9,40.9,78.1,67.7]
+FL  = [78.2,69.8,78.6,77.8,78.2,53.4,74.9,86.8,76.2]
+
+# ── Figure 1: per-film presence F1, three boundary sources ───────────────────
+def fig_presence():
+    x = np.arange(len(FILMS)); w = 0.26
+    fig, ax = plt.subplots(figsize=(11,5))
+    ax.bar(x-w, TE, w, label="track-extent (flood off)", color="#9aa7b4")
+    ax.bar(x,   FG, w, label="flood + grayscale cuts",   color="#e07a5f")
+    ax.bar(x+w, FL, w, label="flood + learned detector (LOO)", color="#3d7ea6")
+    ax.set_ylabel("per-second X-Ray presence F1 (%)")
+    ax.set_title("Actor-presence accuracy by flood-fill boundary source (leave-one-out)")
+    ax.set_xticks(x); ax.set_xticklabels(FILMS, rotation=30, ha="right")
+    ax.set_ylim(0,100); ax.legend(loc="upper left", framealpha=0.9)
+    # annotate the two headline swings
+    ax.annotate("grayscale flood\nBREAKS Scarface", xy=(6, 40.9), xytext=(5.1, 20),
+                fontsize=9, color="#b23", ha="center",
+                arrowprops=dict(arrowstyle="->", color="#b23"))
+    ax.annotate("+37pp", xy=(2+w, 78.6), xytext=(2+w, 90), fontsize=9,
+                color="#3d7ea6", ha="center",
+                arrowprops=dict(arrowstyle="->", color="#3d7ea6"))
+    macro=[np.mean(TE),np.mean(FG),np.mean(FL)]
+    ax.text(0.99,0.02,f"macro: {macro[0]:.1f}% / {macro[1]:.1f}% / {macro[2]:.1f}%",
+            transform=ax.transAxes, ha="right", va="bottom", fontsize=10,
+            bbox=dict(boxstyle="round", fc="#f4f4f4", ec="#ccc"))
+    fig.tight_layout(); fig.savefig(OUT/"scene_presence_by_source.png"); plt.close(fig)
+
+# ── Figure 2: macro presence F1 — the progression ───────────────────────────
+def fig_macro():
+    labels=["track-extent","flood +\ngrayscale","flood +\nlearned (LOO)"]
+    vals=[np.mean(TE),np.mean(FG),np.mean(FL)]
+    fig,ax=plt.subplots(figsize=(6,4.5))
+    bars=ax.bar(labels,vals,color=["#9aa7b4","#e07a5f","#3d7ea6"])
+    for b,v in zip(bars,vals): ax.text(b.get_x()+b.get_width()/2, v+1, f"{v:.1f}%",
+                                       ha="center", fontsize=11, fontweight="bold")
+    ax.set_ylabel("macro presence F1 (%)"); ax.set_ylim(0,90)
+    ax.set_title("Flood-fill boundary source → presence accuracy")
+    fig.tight_layout(); fig.savefig(OUT/"scene_presence_macro.png"); plt.close(fig)
+
+# ── Figure 3: feature/model evolution (boundary-F1 development) ──────────────
+# Two panels, because the development curve and the shipped result are measured
+# at DIFFERENT tolerances and must not be plotted on one axis:
+#   left  — relative feature progress at the strict ±2 s tolerance (how the LSTM
+#           experiments were scored; establishes which features helped)
+#   right — the shipped XGBoost detector at the ±20 s tolerance the pipeline
+#           actually uses and scores at (grayscale vs learned-LOO vs train-all)
+def fig_evolution():
+    fig,(axl,axr)=plt.subplots(1,2,figsize=(11,4.5),gridspec_kw={"width_ratios":[1.15,1]})
+
+    steps=["grayscale\nbaseline","raw-hist\nLSTM","delta\nLSTM","XGBoost\n(delta+debounce)"]
+    dev=[7.2,7.5,10.8,15.2]   # boundary-F1 @±2s during LSTM-era development
+    axl.plot(steps,dev,marker="o",color="#9aa7b4",lw=2,ms=8)
+    for i,v in enumerate(dev): axl.text(i,v+0.4,f"{v:.1f}%",ha="center",fontsize=9)
+    axl.set_ylabel("boundary F1 @±2 s (%)")
+    axl.set_title("Feature progress (strict ±2 s)")
+    axl.set_ylim(0,18)
+
+    # shipped detector at the ±20s tolerance the pipeline uses — real measured
+    # macro numbers: grayscale (xgb_report gray_F1), learned LOO, learned train-all
+    names=["grayscale","learned\n(LOO)","learned\n(train-all)"]
+    f20=[29.8,44.1,72.9]; cols=["#e07a5f","#3d7ea6","#8fb8cf"]
+    bars=axr.bar(names,f20,color=cols)
+    for b,v in zip(bars,f20): axr.text(b.get_x()+b.get_width()/2,v+1.2,f"{v:.1f}%",
+                                       ha="center",fontsize=10,fontweight="bold")
+    axr.set_ylabel("boundary F1 @±20 s (%)")
+    axr.set_title("Shipped detector (±20 s, macro/9 films)")
+    axr.set_ylim(0,80)
+    fig.suptitle("Detector development, and where it landed",fontsize=13)
+    fig.tight_layout(); fig.savefig(OUT/"scene_detector_evolution.png"); plt.close(fig)
+
+import csv as _csv
+
+# ── Figure 4: DE convergence (the 10-knob presence sweep) ────────────────────
+def fig_de():
+    import json
+    rows=[json.loads(l) for l in open("experiments/trajectories/lvface_opencv5_10knob.FINAL.jsonl")]
+    f1=[r["f1"]*100 for r in rows]
+    run_best=np.maximum.accumulate(f1)
+    fig,ax=plt.subplots(figsize=(8,4.5))
+    ax.scatter(range(len(f1)),f1,s=8,alpha=0.35,color="#9aa7b4",label="candidate")
+    ax.plot(run_best,color="#3d7ea6",lw=2,label="best so far")
+    ax.set_xlabel("DE evaluation"); ax.set_ylabel("macro presence F1 (%)")
+    ax.set_title("10-knob presence sweep (Differential Evolution)")
+    ax.legend(loc="lower right"); ax.set_ylim(0, max(f1)+8)
+    ax.text(0.02,0.95,f"optimum {max(f1):.1f}%",transform=ax.transAxes,va="top",
+            fontsize=10,bbox=dict(boxstyle="round",fc="#f4f4f4",ec="#ccc"))
+    fig.tight_layout(); fig.savefig(OUT/"de_search_landscape.png"); plt.close(fig)
+
+# ── Figure 5: calibration curve (similarity → P(match)) ──────────────────────
+def fig_calibration():
+    sims,ps=[],[]
+    with open("experiments/galleries/gallery_LVFace-B_Glint360K.h5.calib_cache.csv") as f:
+        for r in _csv.DictReader(f):
+            sims.append(float(r["similarity"])); ps.append(float(r["p_match"]))
+    fig,ax=plt.subplots(figsize=(6.5,4.5))
+    ax.plot(sims,ps,color="#3d7ea6",lw=2)
+    ax.axhline(0.485,ls="--",color="#e07a5f",lw=1,label="shipped threshold 0.485")
+    ax.set_xlabel("cosine similarity"); ax.set_ylabel("calibrated P(match)")
+    ax.set_title("LVFace-B Glint360K calibration"); ax.set_xlim(-1,1); ax.legend()
+    fig.tight_layout(); fig.savefig(OUT/"calibration_curves.png"); plt.close(fig)
+
+# ── Figure 6: holdout F1 by film (learned detector, LOO) ─────────────────────
+def fig_holdout():
+    order=np.argsort(FL)
+    fig,ax=plt.subplots(figsize=(8,4.5))
+    y=np.arange(len(FILMS))
+    ax.barh(y,[FL[i] for i in order],color="#3d7ea6")
+    ax.set_yticks(y); ax.set_yticklabels([FILMS[i] for i in order])
+    ax.set_xlabel("presence F1 (%), learned detector (LOO)")
+    ax.set_title("Per-film presence F1 — leave-one-out")
+    ax.axvline(np.mean(FL),ls="--",color="#333",lw=1)
+    ax.text(np.mean(FL)+1,0.2,f"macro {np.mean(FL):.1f}%",fontsize=9)
+    for i,idx in enumerate(order): ax.text(FL[idx]+0.5,i,f"{FL[idx]:.0f}",va="center",fontsize=8)
+    ax.set_xlim(0,100)
+    fig.tight_layout(); fig.savefig(OUT/"holdout_f1_by_film.png"); plt.close(fig)
+
+fig_presence(); fig_macro(); fig_evolution(); fig_de(); fig_calibration(); fig_holdout()
+print("wrote:", *(p.name for p in sorted(OUT.glob("*.png"))))

scripts/scene_detector/rematch_frames.py

@@ -0,0 +1,116 @@
+#!/usr/bin/env python3
+"""
+rematch_frames.py — remake each named July frame example against the CURRENT
+pipeline. For a file named <film>_<...>_<actor>.jpg, find a second in this film's
+replay where that actor is drawn in the matching class (FP for *_fpi_*, TP for
+*_tp/perfect*), extract + annotate it, and write it over the doc asset. Reports
+which July examples no longer reproduce (honest — the config/model changed).
+
+Needs the per-film raw replay (experiments/dumps + replay --raw-out already run by
+regen_frame_examples.sh into the scratch predictions). Reads those.
+"""
+from __future__ import annotations
+import json, sys, subprocess, re
+from pathlib import Path
+sys.path.insert(0, "scripts/optimizer"); sys.path.insert(0, "scripts/validation")
+import dump_error_frames as D
+from second_score import load_second_timeline, _match
+
+SP = Path("/tmp/claude-1000/-home-dtourolle-Development-scene-actor-extraction/"
+          "c579f8cf-2974-4cbd-be88-afec68dbbf58/scratchpad")
+ASSETS = Path("docs/assets/images")
+LUT = json.load(open("experiments/file-lut.json"))
+FILMS = json.load(open("experiments/manifests/films_LVFace_opencv5.json"))
+XR = {f["slug"]: f["xray"] for f in FILMS}
+
+# filename → (film slug, actor substring, class). class: "fp" | "tp".
+# actor substring is matched case-insensitively against drawn names.
+JOBS = {
+    "lord_of_war_fpi_reddick.jpg":       ("Lord_of_War", "reddick", "fp"),
+    "lord_of_war_fpi_shumbris.jpg":      ("Lord_of_War", "shumbris", "fp"),
+    "lord_of_war_fpi_reagan_photo.jpg":  ("Lord_of_War", "reagan", "fp"),
+    "lovelace_fpi_sevigny.jpg":          ("Lovelace", "sevigny", "fp"),
+    "lovelace_robert_patrick_fpi.jpg":   ("Lovelace", "patrick", "fp"),
+    "lovelace_perfect_second.jpg":       ("Lovelace", None, "tp"),
+    "lovelace_polygraph_bridged.jpg":    ("Lovelace", None, "tp"),
+    "many_saints_fpi_deschanel.jpg":     ("The_Many_Saints_of_Newark", "deschanel", "fp"),
+    "many_saints_fpi_gardner.jpg":       ("The_Many_Saints_of_Newark", "gardner", "fp"),
+    "many_saints_fpi_yates.jpg":         ("The_Many_Saints_of_Newark", "yates", "fp"),
+    "many_saints_outofcast_fpi.jpg":     ("The_Many_Saints_of_Newark", None, "fp"),
+    "scarface_fpi_alley.jpg":            ("Scarface", "alley", "fp"),
+    "downton_crew_fn.jpg":               ("Downton_Abbey__A_New_Era", None, "tp"),
+    "downton_wedding_couple.jpg":        ("Downton_Abbey__A_New_Era", None, "tp"),
+    "downton_tp_example.jpg":            ("Downton_Abbey__A_New_Era", None, "tp"),
+    "valerian_screen_call.jpg":          ("Valerian_and_the_City_of_a_Thousand_Plan", None, "tp"),
+    "cafe_society_rapid_cut.jpg":        ("Café_Society", None, "tp"),
+    # germar_beats_xray / downton_funeral_19of20 are July-narrative-specific; skip.
+}
+
+
+def gt_keysets(slug):
+    tl, _, _ = load_second_timeline(XR[slug])
+    return tl
+
+
+def main():
+    made, missing = [], []
+    for fname, (slug, actor, cls) in JOBS.items():
+        raw = SP / f"{slug}_raw.jsonl"
+        if not raw.exists():
+            missing.append((fname, "no raw replay")); continue
+        tl = gt_keysets(slug)
+        best = None  # (t, actor_dict, fp_keys)
+        for line in open(raw):
+            d = json.loads(line)
+            t = int(d["timestamp_sec"])
+            drawn = [a for a in d.get("visible_actors", []) if a.get("actor_idx", -1) >= 0]
+            if not drawn:
+                continue
+            gt = tl.get(t, [])
+            fp_keys = {D._name_key(a["name"]) for a in drawn
+                       if not any(D._name_key(a["name"]) in g for g in gt)}
+            for a in drawn:
+                nk = D._name_key(a["name"]); is_fp = nk in fp_keys
+                if actor and actor not in a["name"].lower():
+                    continue
+                match = (is_fp if cls == "fp" else not is_fp)
+                if not match:
+                    continue
+                # prefer high similarity + a clean single-subject frame
+                score = a["similarity"] - 0.05*len(drawn)
+                if best is None or score > best[3]:
+                    best = (t, d, fp_keys, score)
+        if best is None:
+            missing.append((fname, f"no current {cls} for {actor or 'any'}")); continue
+        t, d, fp_keys, _ = best
+        # FN names at t: X-Ray scene cast whose keyset matches no drawn face.
+        gt = tl.get(t, [])
+        drawn_keys = [set(D._name_key(a["name"]).replace("name:", "") for _ in [0])
+                      for a in d.get("visible_actors", []) if a.get("actor_idx", -1) >= 0]
+        drawn_ks = [D._name_key(a["name"]) for a in d.get("visible_actors", [])
+                    if a.get("actor_idx", -1) >= 0]
+        fn_names = []
+        for ga in gt:
+            if not any(dk in ga for dk in drawn_ks):
+                readable = sorted(x for x in ga
+                                  if not x.startswith("imdb:") and not x.startswith("tmdb:")
+                                  and not x.startswith("jf:"))
+                if readable:
+                    fn_names.append(readable[0])
+        out = ASSETS / fname
+        try:
+            D.extract_frame(LUT[slug], t, out)
+            D.draw_annotations(out, d["visible_actors"], fp_keys=fp_keys,
+                               fn_names=fn_names)
+            made.append((fname, slug, t))
+        except subprocess.CalledProcessError:
+            missing.append((fname, "ffmpeg failed"))
+
+    print("=== remade ===")
+    for f, s, t in made: print(f"  {f}  ({s} t={t}s)")
+    print("=== no current equivalent (left as-is / flag in doc) ===")
+    for f, why in missing: print(f"  {f}  — {why}")
+
+
+if __name__ == "__main__":
+    main()

scripts/scene_detector/run_de_audio_cutter.py

@@ -0,0 +1,56 @@
+#!/usr/bin/env python3
+"""Standalone DE-optimised AUDIO scene cutter: tune a matched-filter ramp on the
+audio log-PSD to maximise X-Ray boundary F1. No neural net. Holdout films are
+never seen in training. Writes the tuned filter + held-out performance."""
+import sys, json, os
+import numpy as np
+sys.path.insert(0, "scripts/scene_detector")
+from de_ramp import load_series, xray_bounds, ramp_kernel, response, boundary_f1
+from scipy.optimize import differential_evolution
+
+MANIFEST = "experiments/manifests/films_LVFace_opencv5.json"
+AUDIO = "experiments/dumps/audio_features"
+HOLDOUT = {"Scarface", "Sound_of_Metal", "Valerian_and_the_City_of_a_Thousand_Plan"}
+OUT = "experiments/results/scene_boundary/de_audio_cutter.json"
+
+films = json.load(open(MANIFEST))
+train = [f for f in films if f["slug"] not in HOLDOUT]
+val = [f for f in films if f["slug"] in HOLDOUT]
+tr = [(load_series(f["dump"], AUDIO, "audio")[0], xray_bounds(f["xray"])) for f in train]
+va = [(f["slug"], load_series(f["dump"], AUDIO, "audio")[0], xray_bounds(f["xray"])) for f in val]
+print(f"DE AUDIO cutter: {len(tr)} train, holdout {sorted(HOLDOUT)}", flush=True)
+
+def neg_f1(x):
+    H = int(round(x[0])); gamma = x[1]; dead = int(round(x[2])); pct = x[3]
+    if H < 1 or dead >= H: return 0.0
+    w = ramp_kernel(H, gamma, dead)
+    return -float(np.mean([boundary_f1(response(s, w, H), b, pct) for s, b in tr]))
+
+evals = [0]
+def cb(xk, convergence):
+    evals[0] += 1
+    print(f"[de-audio] gen {evals[0]}  convergence={convergence:.3f}", flush=True)
+
+res = differential_evolution(neg_f1, [(1, 10), (0.3, 3.0), (0, 4), (80, 98)],
+                             seed=0, popsize=12, maxiter=25, tol=1e-4,
+                             polish=False, callback=cb)
+H = int(round(res.x[0])); gamma = float(res.x[1]); dead = int(round(res.x[2])); pct = float(res.x[3])
+print(f"\n=== DE-OPTIMISED AUDIO SCENE CUTTER ===", flush=True)
+print(f"tuned ramp: H={H}s gamma={gamma:.2f} dead={dead}s threshold_pct={pct:.0f}", flush=True)
+print(f"train boundary-F1: {-res.fun*100:.1f}%\n", flush=True)
+print("held-out (audio-only, P/R/F1 ±2s):", flush=True)
+w = ramp_kernel(H, gamma, dead)
+rep = {"H": H, "gamma": gamma, "dead": dead, "pct": pct,
+       "train_f1": float(-res.fun), "holdout": sorted(HOLDOUT), "films": {}}
+for slug, s, b in va:
+    r = response(s, w, H); thr = np.percentile(r, pct); pred = np.where(r > thr)[0]
+    bidx = [int(x) for x in b if int(x) < len(r)]
+    tp_p = sum(any(abs(p-i) <= 2 for i in bidx) for p in pred)
+    tp_t = sum(any(abs(p-i) <= 2 for p in pred) for i in bidx)
+    P = tp_p/max(len(pred), 1); R = tp_t/max(len(bidx), 1); F = 2*P*R/(P+R) if P+R else 0
+    rep["films"][slug] = {"P": P, "R": R, "F1": F, "n_pred": len(pred), "n_true": len(bidx)}
+    print(f"  {slug[:26]:26s} P={P*100:4.0f}% R={R*100:4.0f}% F1={F*100:4.0f}%  "
+          f"({len(pred)} preds/{len(bidx)} true)", flush=True)
+os.makedirs(os.path.dirname(OUT), exist_ok=True)
+json.dump(rep, open(OUT, "w"), indent=2)
+print(f"\nsaved → {OUT}", flush=True)

scripts/scene_detector/train_scene_boundary.py

@@ -0,0 +1,356 @@
+#!/usr/bin/env python3
+"""
+train_scene_boundary.py — learn a scene-boundary detector from per-frame RGB
+histograms (video tower) and per-second audio log-PSD (audio tower), against
+Amazon X-Ray scene boundaries.
+
+Motivation: the shipped grayscale histogram-correlation cut detector is blind on
+low-contrast grades — on Scarface it fired ONCE in 10,204 frames, so flood-fill
+presence (which snaps to detected boundaries) floods every actor across the whole
+film (P=26%). X-Ray ships real scene boundaries (scenes.csv); the dumps carry a
+per-frame RGB histogram (frames/rgb_hist), and extract_audio_features.py provides
+a per-second audio log-PSD. This learns a per-second boundary probability.
+
+TWO-TOWER, ABLATABLE. We do NOT assume audio helps video — we measure it. Each
+modality has its own encoder+BiLSTM; --modality selects video / audio / fused
+(both towers concatenated before a shared head). The script reports all three
+arms on the held-out films so the ablation decides whether audio supports video.
+
+Video features per second: rgb_hist (96) + L1 deltas to t-1,t-2,t+1 + per-channel
+correlation to t-1. Audio features: the log-PSD row (+ its L1 delta to t-1).
+Label: 1 if an X-Ray scene starts within ±TOL_SEC of t.
+
+Usage:
+  python scripts/scene_detector/train_scene_boundary.py \
+      --manifest experiments/manifests/films_LVFace_opencv5.json \
+      --audio-dir experiments/dumps/audio_features \
+      --holdout Scarface Sound_of_Metal \
+      --modality all --out experiments/results/scene_boundary
+"""
+from __future__ import annotations
+import argparse, csv, json, sys
+from pathlib import Path
+
+import h5py
+import numpy as np
+import torch
+import torch.nn as nn
+
+TOL_SEC = 2.0
+BINS = 32          # per channel, matches embedding_dump_node.hpp kHistBins
+RAMP_SCALES = [2, 4, 6, 8, 10]   # multi-scale matched-filter half-widths (seconds)
+SCENE_TAU = 205.0   # corpus mean X-Ray scene length (central-60min); debounce scale
+
+
+def debounce_phase(delta_signal: np.ndarray, tau: float = SCENE_TAU,
+                   peak_pct: float = 90.0) -> np.ndarray:
+    """A scene-length-scaled 'how overdue is a boundary' feature, [T,2].
+
+    Encodes the prior that scenes don't restart moments apart. From the strong
+    peaks of a change signal (the presumed boundaries so far), track time since
+    the last peak and turn it into:
+      phase = min(1, dt/tau)   — 0 just after a boundary (suppress), 1 when a new
+                                 one is overdue (permit), rising over ~one mean
+                                 scene length (tau).
+      decay = exp(-dt/tau)     — the complementary refractory (high right after,
+                                 decaying away). Two views of the same clock so
+                                 the LSTM can use whichever helps.
+    Reference peaks come from the change signal itself (not the model's own
+    output), so the feature is static and causal-ish (uses only |Δ| already in
+    the sequence)."""
+    T = len(delta_signal)
+    thr = np.percentile(delta_signal, peak_pct)
+    # Vectorised time-since-last-peak: index of the most recent peak at or before
+    # each t (running max of peak indices), then dt = t - that index.
+    idx = np.arange(T)
+    peak_idx = np.where(delta_signal > thr, idx, -1)
+    last = np.maximum.accumulate(peak_idx)          # most recent peak index ≤ t
+    dt = (idx - last).astype(np.float32)
+    dt[last < 0] = tau                              # before the first peak: treat as "overdue"
+    phase = np.minimum(1.0, dt / tau)
+    decay = np.exp(-dt / tau)
+    return np.stack([phase, decay], 1).astype(np.float32)
+
+
+def ramp_bank(series: np.ndarray) -> np.ndarray:
+    """Antisymmetric matched-filter responses at RAMP_SCALES → [T, len(scales)].
+
+    A scene boundary is a step in the feature series; a signed ramp kernel
+    convolved with it responds at the transition and ~0 inside a stable scene.
+    Different films' boundaries peak at different scales (measured: sharp cuts at
+    H=2s, gradual shifts wider), so we hand the model the whole bank and let it
+    weight the scales rather than committing to one width."""
+    # Vectorised: the ramp response at t is || sum_l w(l)·series[t+l] ||, i.e. a
+    # 1D correlation of the kernel with each feature bin, then an L2 over bins. Do
+    # it as one convolution per bin (np.convolve, 'same') instead of the per-frame
+    # Python loop — ~100x faster, which matters at ~60k frames × 9 films.
+    T, D = series.shape
+    out = np.zeros((T, len(RAMP_SCALES)), np.float32)
+    for k, H in enumerate(RAMP_SCALES):
+        lags = np.arange(-H, H + 1)
+        w = (np.sign(lags) * (np.abs(lags) / max(H, 1))).astype(np.float64)
+        # correlation = convolution with the reversed kernel; ramp is antisym so
+        # reversing negates it — sign folds into the L2 norm, so either is fine.
+        acc = np.zeros((T, D))
+        for d in range(D):
+            acc[:, d] = np.convolve(series[:, d], w[::-1], mode="same")
+        out[:, k] = np.linalg.norm(acc, axis=1)
+    return out
+
+
+# ── data ──────────────────────────────────────────────────────────────────────
+
+def load_xray_boundaries(xray_dir: str) -> list[float]:
+    starts = []
+    with open(Path(xray_dir) / "scenes.csv", newline="") as f:
+        for r in csv.DictReader(f):
+            s = float(r["start"]) / 1000.0
+            if s > 0.5:
+                starts.append(s)
+    return sorted(starts)
+
+
+def _znorm(s):
+    return (s - s.mean(0)) / (s.std(0) + 1e-6)
+
+
+def video_features(hist: np.ndarray) -> np.ndarray:
+    """DELTA-FORWARD video features.
+
+    Measured on the corpus: the raw 96-bin histogram barely separates X-Ray
+    boundaries (~1.4x boundary response) — it encodes what the frame *looks like*,
+    not that it *changed* — while the symmetric histogram delta |hist(t+k)-hist(t-k)|
+    separates them strongly (|Δ 1s| ~4-5x). Feeding 96 dims of raw content
+    diluted the LSTM, so we drop it and lead with multi-scale symmetric deltas,
+    keeping only a compact per-channel-energy summary as context.
+
+    Channels:
+      - symmetric L1 delta |hist(t+k) - hist(t-k)| at k=1,2,4,8s (the boundary cue)
+      - per-channel correlation to the previous second (3)
+      - the multi-scale antisymmetric ramp bank (regional step response)
+      - 3-D per-channel total energy (compact content context, not the full hist)
+    """
+    T = hist.shape[0]
+    def sym_delta(k):
+        fwd = np.roll(hist, -k, 0); fwd[-k:] = hist[-1]
+        bwd = np.roll(hist,  k, 0); bwd[:k] = hist[0]
+        return np.abs(fwd - bwd).sum(1, keepdims=True)
+    deltas = np.concatenate([sym_delta(k) for k in (1, 2, 4, 8)], 1)
+    p1 = np.roll(hist, 1, 0); p1[0] = hist[0]
+    corr = np.zeros((T, 3), np.float32)
+    for c in range(3):
+        a = hist[:, c*BINS:(c+1)*BINS]; b = p1[:, c*BINS:(c+1)*BINS]
+        am, bm = a - a.mean(1, keepdims=True), b - b.mean(1, keepdims=True)
+        corr[:, c] = (am*bm).sum(1) / (np.sqrt((am*am).sum(1)*(bm*bm).sum(1))+1e-9)
+    energy = np.stack([hist[:, c*BINS:(c+1)*BINS].sum(1) for c in range(3)], 1)
+    # scene-length-scaled debounce: 'how overdue is a boundary', from the |Δ1s|
+    # change signal. Encodes that scenes don't restart moments apart (tau=205s).
+    debounce = debounce_phase(deltas[:, 0])
+    return np.concatenate([deltas, corr, ramp_bank(_znorm(hist)), energy, debounce],
+                          1).astype(np.float32)
+
+
+def audio_features(psd: np.ndarray) -> np.ndarray:
+    """DELTA-FORWARD audio features (same principle as video).
+
+    The raw log-PSD is spectral CONTENT (what the audio sounds like), which the DE
+    cutter showed barely localizes X-Ray boundaries. Lead with the CHANGE in the
+    spectrum — symmetric PSD deltas |psd(t+k)-psd(t-k)| at several scales — plus
+    the ramp bank and a compact total-energy summary; drop the full raw PSD.
+    """
+    def sym_delta(k):
+        fwd = np.roll(psd, -k, 0); fwd[-k:] = psd[-1]
+        bwd = np.roll(psd,  k, 0); bwd[:k] = psd[0]
+        return np.abs(fwd - bwd).sum(1, keepdims=True)
+    deltas = np.concatenate([sym_delta(k) for k in (1, 2, 4, 8)], 1)
+    energy = psd.sum(1, keepdims=True)
+    debounce = debounce_phase(deltas[:, 0])
+    return np.concatenate([deltas, ramp_bank(_znorm(psd)), energy, debounce],
+                          1).astype(np.float32)
+
+
+def build_film(dump: str, xray_dir: str, audio_dir: str | None):
+    with h5py.File(dump, "r") as f:
+        if "frames/rgb_hist" not in f:
+            raise SystemExit(f"{dump}: no frames/rgb_hist — re-dump with the "
+                             f"RGB-histogram build of dump_embeddings.")
+        hist = f["frames/rgb_hist"][:].astype(np.float32)
+        ts = f["frames/timestamp_sec"][:]
+        is_cut = f["frames/is_cut"][:].astype(np.int64)
+    V = video_features(hist)
+    A = None
+    if audio_dir:
+        slug = Path(dump).stem.replace("dump_", "")
+        ap = Path(audio_dir) / f"{slug}.npz"
+        if ap.exists():
+            z = np.load(ap); af = z["feat"]
+            # align audio (per-second) to the video frame grid by index; pad/truncate
+            T = len(ts); B = af.shape[1]
+            aligned = np.zeros((T, B), np.float32)
+            m = min(T, len(af)); aligned[:m] = af[:m]
+            A = audio_features(aligned)
+    y = np.zeros(len(ts), np.float32)
+    for b in load_xray_boundaries(xray_dir):
+        y[np.abs(ts - b) <= TOL_SEC] = 1.0
+    return V, A, y, is_cut, ts
+
+
+# ── model ─────────────────────────────────────────────────────────────────────
+
+class Tower(nn.Module):
+    """Per-second encoder → BiLSTM → per-timestep embedding."""
+    def __init__(self, in_dim, hidden=64, out=64):
+        super().__init__()
+        self.enc = nn.Sequential(nn.Linear(in_dim, hidden), nn.ReLU())
+        self.lstm = nn.LSTM(hidden, out, batch_first=True, bidirectional=True)
+    def forward(self, x):
+        h, _ = self.lstm(self.enc(x))
+        return h                       # [B,T,2*out]
+
+
+class BoundaryNet(nn.Module):
+    def __init__(self, v_dim, a_dim, modality):
+        super().__init__()
+        self.modality = modality
+        feat = 0
+        if modality in ("video", "fused"):
+            self.vtower = Tower(v_dim); feat += 128
+        if modality in ("audio", "fused"):
+            self.atower = Tower(a_dim); feat += 128
+        self.head = nn.Sequential(nn.Linear(feat, 32), nn.ReLU(), nn.Linear(32, 1))
+    def forward(self, v, a):
+        parts = []
+        if self.modality in ("video", "fused"): parts.append(self.vtower(v))
+        if self.modality in ("audio", "fused"): parts.append(self.atower(a))
+        return self.head(torch.cat(parts, -1)).squeeze(-1)
+
+
+def nms_peaks(prob, thr=0.5, min_gap=5):
+    """Collapse each run of adjacent above-threshold seconds to its single peak.
+    Without this, a model that fires 5 consecutive seconds around one true
+    boundary is scored as 1 TP + 4 FP — an aggregation artifact, not an error."""
+    cand = np.where(prob > thr)[0]
+    if len(cand) == 0:
+        return []
+    peaks, group = [], [cand[0]]
+    for c in cand[1:]:
+        if c - group[-1] <= min_gap:
+            group.append(c)
+        else:
+            peaks.append(group[int(np.argmax(prob[group]))]); group = [c]
+    peaks.append(group[int(np.argmax(prob[group]))])
+    return peaks
+
+
+def prf(prob_or_pred, y, tol=2, thr=0.5):
+    """Boundary P/R/F1 with NMS peak aggregation. Accepts a probability series
+    (model output) or a 0/1 array (is_cut baseline); NMS collapses each run of
+    above-threshold seconds to one peak either way."""
+    P = np.array(nms_peaks(np.asarray(prob_or_pred, float), thr=thr))
+    T = np.where(y > 0.5)[0]
+    if len(P) == 0 or len(T) == 0: return 0., 0., 0.
+    tp_p = sum(any(abs(p-t) <= tol for t in T) for p in P)
+    tp_t = sum(any(abs(p-t) <= tol for p in P) for t in T)
+    pr, rc = tp_p/len(P), tp_t/len(T)
+    return pr, rc, (2*pr*rc/(pr+rc) if pr+rc else 0.)
+
+
+def train_arm(modality, tr, va, v_dim, a_dim, vmu, vsd, amu, asd, epochs, dev):
+    model = BoundaryNet(v_dim, a_dim, modality).to(dev)
+    opt = torch.optim.Adam(model.parameters(), lr=1e-3, weight_decay=1e-5)
+    pos = sum((y > .5).sum() for *_, y, _, _ in tr)
+    neg = sum((y <= .5).sum() for *_, y, _, _ in tr)
+    lossf = nn.BCEWithLogitsLoss(pos_weight=torch.tensor([neg/max(pos,1)], device=dev))
+    def vt(V): return torch.tensor((V-vmu)/vsd, dtype=torch.float32, device=dev).unsqueeze(0)
+    def at(A): return torch.tensor((A-amu)/asd, dtype=torch.float32, device=dev).unsqueeze(0)
+    for ep in range(epochs):
+        model.train()
+        for V, A, y, _, _ in tr:
+            opt.zero_grad()
+            logit = model(vt(V), at(A) if A is not None else None)
+            loss = lossf(logit, torch.tensor(y, device=dev).unsqueeze(0))
+            loss.backward(); opt.step()
+    model.eval(); rows = {}
+    with torch.no_grad():
+        for slug, V, A, y, is_cut, ts in va:
+            prob = torch.sigmoid(model(vt(V), at(A) if A is not None else None))[0].cpu().numpy()
+            rows[slug] = prf(prob, y)   # raw prob → NMS picks peaks by height
+    return model, rows
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--manifest", required=True)
+    ap.add_argument("--audio-dir", default="experiments/dumps/audio_features")
+    ap.add_argument("--holdout", nargs="+", default=["Scarface", "Sound_of_Metal"])
+    ap.add_argument("--modality", choices=["video","audio","fused","all"], default="all")
+    ap.add_argument("--out", default="experiments/results/scene_boundary")
+    ap.add_argument("--epochs", type=int, default=250)
+    ap.add_argument("--seed", type=int, default=0)
+    args = ap.parse_args()
+    torch.manual_seed(args.seed); np.random.seed(args.seed)
+
+    films = json.load(open(args.manifest))
+    def load(rows):
+        out = []
+        for f in rows:
+            V, A, y, is_cut, ts = build_film(f["dump"], f["xray"], args.audio_dir)
+            out.append((f["slug"], V, A, y, is_cut, ts))
+        return out
+    tr = load([f for f in films if f["slug"] not in args.holdout])
+    va = load([f for f in films if f["slug"] in args.holdout])
+    has_audio = all(t[2] is not None for t in tr+va)
+    print(f"[scene] train {len(tr)} / holdout {args.holdout}; audio={'yes' if has_audio else 'MISSING'}",
+          file=sys.stderr)
+
+    allV = np.concatenate([t[1] for t in tr], 0)
+    vmu, vsd = allV.mean(0), allV.std(0)+1e-6; v_dim = allV.shape[1]
+    if has_audio:
+        allA = np.concatenate([t[2] for t in tr], 0)
+        amu, asd = allA.mean(0), allA.std(0)+1e-6; a_dim = allA.shape[1]
+    else:
+        amu = asd = None; a_dim = 1
+
+    # strip index tuples for train_arm (expects V,A,y,is_cut,ts)
+    trA = [(t[1],t[2],t[3],t[4],t[5]) for t in tr]
+    dev = "cuda" if torch.cuda.is_available() else "cpu"
+    modes = ["video","audio","fused"] if args.modality=="all" else [args.modality]
+    if not has_audio: modes = [m for m in modes if m == "video"] or ["video"]
+
+    # grayscale-0.70 baseline (is_cut) on holdout
+    print("\n=== held-out scene-boundary detection (P/R/F1, ±2s) ===")
+    print(f"{'film':26s} " + "  ".join(f"{m:>16s}" for m in modes) + f"  {'grayscale-0.70':>16s}")
+    Path(args.out).mkdir(parents=True, exist_ok=True)
+    results = {m: train_arm(m, trA, va, v_dim, a_dim, vmu, vsd, amu, asd, args.epochs, dev)
+               for m in modes}
+    report = {"holdout": args.holdout, "tol_sec": TOL_SEC, "modalities": {}, "films": {}}
+    for slug, V, A, y, is_cut, ts in va:
+        cells = []
+        for m in modes:
+            p,r,f = results[m][1][slug]
+            cells.append(f"{p*100:4.0f}/{r*100:4.0f}/{f*100:4.0f}")
+            report["films"].setdefault(slug, {})[m] = {"P":p,"R":r,"F1":f}
+        bp,br,bf = prf(is_cut, y)
+        report["films"].setdefault(slug, {})["grayscale"] = {"P":bp,"R":br,"F1":bf}
+        print(f"{slug:26s} " + "  ".join(f"{c:>16s}" for c in cells) +
+              f"  {bp*100:4.0f}/{br*100:4.0f}/{bf*100:4.0f}")
+    # macro-mean F1 per modality across holdout
+    print("\nmacro-mean holdout F1:")
+    for m in modes:
+        mf = np.mean([results[m][1][s][2] for s,*_ in va])
+        report["modalities"][m] = float(mf)
+        print(f"  {m:8s} {mf*100:.1f}%")
+    bf = np.mean([prf(t[4], t[3])[2] for t in va])
+    report["modalities"]["grayscale"] = float(bf)
+    print(f"  {'grayscale':8s} {bf*100:.1f}%")
+    # save the best arm
+    best = max(modes, key=lambda m: report["modalities"][m])
+    torch.save({"state": results[best][0].state_dict(), "modality": best,
+                "vmu":vmu,"vsd":vsd,"amu":amu,"asd":asd,"v_dim":v_dim,"a_dim":a_dim},
+               Path(args.out)/"boundary_net.pt")
+    json.dump(report, open(Path(args.out)/"report.json","w"), indent=2)
+    print(f"\n[scene] best={best}; model+report → {args.out}", file=sys.stderr)
+
+
+if __name__ == "__main__":
+    main()

scripts/scene_detector/train_xgb_boundary.py

@@ -0,0 +1,155 @@
+#!/usr/bin/env python3
+"""
+train_xgb_boundary.py — SHIPPED scene-boundary detector.
+
+An XGBoost regressor over a ±WIN-second window of delta features predicts a soft
+Gaussian proximity-to-boundary target; a per-film KNEE threshold on the predicted
+peak heights selects the boundaries (self-calibrates the count without a magic
+rate). Evaluated with NMS + P/R/F1 at ±20 s tolerance (X-Ray scenes are ~170 s,
+so ±20 s placement is what flood-fill actually needs).
+
+Why this shape (all measured, see docs/scene-detector):
+  - DELTA features, not raw histogram/PSD: the raw content dilutes; |Δ| separates
+    boundaries 4-5x. Audio is weak but included (XGBoost ignores what it can't use).
+  - SOFT target exp(-(d/σ)²), σ=10s: a near-miss is trained as near-correct, not a
+    hard negative. Regression → smooth score surface → NMS peaks.
+  - KNEE threshold per film: peak-height curve has a knee where real boundaries
+    give way to noise; picking it matches the true scene count without a global
+    threshold that's wrong for every grade.
+  - Café Society + Scarface (low-contrast grades) MUST be in training; held out,
+    the model can't generalize to them. The shipped model trains on ALL 9.
+
+Honest generalization: leave-one-out CV ≈ 26% F1 @±10s / ~34% @±20s. The shipped
+all-9 model is what deployment uses (max grade coverage); LOO is the number to
+quote for a brand-new film.
+
+Usage (train on all 9 + save shipped model):
+  .venv-rocm/bin/python scripts/scene_detector/train_xgb_boundary.py --train-all
+Usage (held-out eval):
+  ... --holdout Sound_of_Metal The_Many_Saints_of_Newark Valerian_...
+"""
+from __future__ import annotations
+import argparse, json, sys
+from pathlib import Path
+import numpy as np
+import h5py
+
+sys.path.insert(0, "scripts/scene_detector")
+from train_scene_boundary import nms_peaks, load_xray_boundaries, SCENE_TAU, TOL_SEC
+from train_scene_boundary import video_features, audio_features, build_film
+from scipy.signal import find_peaks
+import xgboost as xgb
+
+WIN = 3          # ±WIN-second context window
+SIGMA = 10.0     # soft-target Gaussian width (seconds)
+
+
+def per_second_matrix(dump, xray, audio_dir, win=None):
+    """Windowed delta features + debounce clock → (X[T,F], y_binary[T], is_cut[T])."""
+    V, A, y, is_cut, ts = build_film(dump, xray, audio_dir)
+    base = np.concatenate([V] + ([A] if A is not None else []), 1)
+    T, d = base.shape
+    sig = V[:, 0]
+    thr = np.percentile(sig, 90)
+    idx = np.arange(T); peak = np.where(sig > thr, idx, -1)
+    last = np.maximum.accumulate(peak)
+    dt = (idx - last).astype(np.float32); dt[last < 0] = SCENE_TAU
+    clock = np.stack([dt, np.minimum(1, dt/SCENE_TAU), np.exp(-dt/SCENE_TAU)], 1)
+    W = WIN if win is None else win
+    padded = np.pad(base, ((W, W), (0, 0)), mode="edge")
+    wf = np.concatenate([padded[i:i+T] for i in range(2*W+1)], 1)
+    return np.concatenate([wf, clock], 1).astype(np.float32), y, is_cut
+
+
+def soft_target(dump, xray):
+    ts = h5py.File(dump)["frames/timestamp_sec"][:]
+    b = np.array(load_xray_boundaries(xray))
+    y = np.zeros(len(ts), np.float32)
+    if len(b):
+        for i, t in enumerate(ts):
+            y[i] = np.exp(-((np.min(np.abs(b - t)))/SIGMA)**2)
+    return y
+
+
+def knee_boundaries(prob, min_gap=5):
+    """Per-film knee threshold on peak heights → selected peak indices.
+
+    Peaks sorted by height form a convex-decreasing curve; the knee (max drop
+    below the endpoints chord) is where real boundaries give way to noise. Returns
+    the timestamps (indices) of peaks at or above the knee height."""
+    pk, _ = find_peaks(prob, distance=min_gap)
+    if len(pk) < 5:
+        return list(pk)
+    heights = np.sort(prob[pk])[::-1]
+    n = len(heights); x = np.arange(n)/(n-1); yv = heights/(heights[0]+1e-9)
+    chord = yv[0] + (yv[-1]-yv[0])*x
+    k = int(np.argmax(chord - yv))
+    thr = heights[k]
+    return [int(i) for i in pk if prob[i] >= thr]
+
+
+def train(films, audio_dir):
+    X = np.concatenate([per_second_matrix(f["dump"], f["xray"], audio_dir)[0] for f in films])
+    y = np.concatenate([soft_target(f["dump"], f["xray"]) for f in films])
+    reg = xgb.XGBRegressor(n_estimators=400, max_depth=5, learning_rate=0.05,
+                           subsample=0.8, colsample_bytree=0.8,
+                           objective="reg:squarederror", n_jobs=8, tree_method="hist")
+    reg.fit(X, y)
+    return reg
+
+
+def prf(peaks, Tset, tol=20):
+    if not peaks or len(Tset) == 0:
+        return 0., 0., 0., 0, 0, len(Tset)
+    tp_p = sum(any(abs(p-t) <= tol for t in Tset) for p in peaks)
+    tp_t = sum(any(abs(p-t) <= tol for p in peaks) for t in Tset)
+    P = tp_p/len(peaks); R = tp_t/len(Tset)
+    return (P, R, (2*P*R/(P+R) if P+R else 0.),
+            tp_p, len(peaks)-tp_p, len(Tset)-tp_t)
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--manifest", default="experiments/manifests/films_LVFace_opencv5.json")
+    ap.add_argument("--audio-dir", default="experiments/dumps/audio_features")
+    ap.add_argument("--holdout", nargs="*", default=[])
+    ap.add_argument("--train-all", action="store_true", help="train on all 9 + save shipped model")
+    ap.add_argument("--tol", type=int, default=20)
+    ap.add_argument("--out", default="experiments/results/scene_boundary")
+    args = ap.parse_args()
+    films = json.load(open(args.manifest))
+    Path(args.out).mkdir(parents=True, exist_ok=True)
+
+    tr = films if args.train_all else [f for f in films if f["slug"] not in args.holdout]
+    reg = train(tr, args.audio_dir)
+    print(f"[xgb] trained on {len(tr)} films", file=sys.stderr)
+
+    ev = films if args.train_all else [f for f in films if f["slug"] in args.holdout]
+    tag = "TRAIN-FIT (all 9)" if args.train_all else "HELD-OUT"
+    print(f"\n=== {tag} boundary detection (knee, NMS, ±{args.tol}s) ===")
+    print(f"{'film':26s} {'TP':>4}{'FP':>5}{'FN':>5} {'P':>5}{'R':>5}{'F1':>5}  {'gray F1':>7}")
+    rep = {"win": WIN, "sigma": SIGMA, "tol": args.tol, "train_all": args.train_all,
+           "holdout": args.holdout, "films": {}}
+    f1s, gf1s = [], []
+    for f in ev:
+        X, yb, ic = per_second_matrix(f["dump"], f["xray"], args.audio_dir)
+        prob = np.clip(reg.predict(X), 0, 1)
+        peaks = knee_boundaries(prob)
+        Tset = np.where(yb > 0.5)[0]
+        P, R, F, tp, fp, fn = prf(peaks, Tset, args.tol)
+        gpk = nms_peaks(ic.astype(float)); _, _, gF, *_ = prf(gpk, Tset, args.tol)
+        f1s.append(F); gf1s.append(gF)
+        rep["films"][f["slug"]] = {"TP": tp, "FP": fp, "FN": fn, "P": P, "R": R, "F1": F,
+                                   "n_pred": len(peaks), "n_true": len(Tset), "gray_F1": gF}
+        print(f"{f['slug'][:26]:26s} {tp:>4}{fp:>5}{fn:>5} {P*100:4.0f}%{R*100:4.0f}%"
+              f"{F*100:4.0f}%  {gF*100:5.0f}%")
+    print(f"\nmacro-F1: detector {np.mean(f1s)*100:.1f}%   grayscale {np.mean(gf1s)*100:.1f}%")
+    rep["macro_f1"] = {"detector": float(np.mean(f1s)), "grayscale": float(np.mean(gf1s))}
+    if args.train_all:
+        reg.save_model(str(Path(args.out) / "xgb_boundary_shipped.json"))
+        print(f"[xgb] shipped model → {args.out}/xgb_boundary_shipped.json", file=sys.stderr)
+    json.dump(rep, open(Path(args.out) / "xgb_report.json", "w"), indent=2)
+
+
+if __name__ == "__main__":
+    main()

scripts/scene_detector/train_xgb_cpp.py

@@ -0,0 +1,81 @@
+#!/usr/bin/env python3
+"""
+train_xgb_cpp.py — train the scene-boundary XGBoost on the C++-EXTRACTED feature
+matrices (experiments/dumps/cpp_features/<slug>.h5, written by scene_features_dump).
+
+This is the parity-by-construction path: the model is fit on exactly the features
+the C++ XGBSceneBoundary produces at inference, so C++ boundaries match by
+construction — no numpy-vs-C++ feature drift to chase. Same soft Gaussian target,
+knee threshold, and ±20s eval as train_xgb_boundary.py.
+
+Usage (train all 9 + save shipped model):
+  .venv-rocm/bin/python scripts/scene_detector/train_xgb_cpp.py --train-all
+"""
+from __future__ import annotations
+import argparse, json, sys
+from pathlib import Path
+import numpy as np, h5py
+sys.path.insert(0, "scripts/scene_detector")
+from train_scene_boundary import load_xray_boundaries, nms_peaks
+from train_xgb_boundary import knee_boundaries, prf, SIGMA
+import xgboost as xgb
+
+CPP_DIR = "experiments/dumps/cpp_features"
+
+
+def load(slug, xray):
+    with h5py.File(f"{CPP_DIR}/{slug}.h5") as f:
+        X = f["features"][:].astype(np.float32)
+        ts = f["timestamp_sec"][:]
+    b = np.array(load_xray_boundaries(xray))
+    y = np.zeros(len(ts), np.float32)
+    if len(b):
+        for i, t in enumerate(ts):
+            y[i] = np.exp(-((np.min(np.abs(b - t)))/SIGMA)**2)
+    yb = np.zeros(len(ts), np.float32)
+    for bb in b:
+        yb[np.abs(ts - bb) <= 2.0] = 1.0
+    return X, y, yb, ts
+
+
+def main():
+    ap = argparse.ArgumentParser()
+    ap.add_argument("--manifest", default="experiments/manifests/films_LVFace_opencv5.json")
+    ap.add_argument("--holdout", nargs="*", default=[])
+    ap.add_argument("--train-all", action="store_true")
+    ap.add_argument("--tol", type=int, default=20)
+    ap.add_argument("--out", default="experiments/results/scene_boundary")
+    args = ap.parse_args()
+    films = json.load(open(args.manifest))
+    Path(args.out).mkdir(parents=True, exist_ok=True)
+
+    tr = films if args.train_all else [f for f in films if f["slug"] not in args.holdout]
+    Xtr = np.concatenate([load(f["slug"], f["xray"])[0] for f in tr])
+    ytr = np.concatenate([load(f["slug"], f["xray"])[1] for f in tr])
+    reg = xgb.XGBRegressor(n_estimators=400, max_depth=5, learning_rate=0.05,
+                           subsample=0.8, colsample_bytree=0.8,
+                           objective="reg:squarederror", n_jobs=8, tree_method="hist")
+    reg.fit(Xtr, ytr)
+    print(f"[xgb-cpp] trained on {len(tr)} films", file=sys.stderr)
+
+    ev = films if args.train_all else [f for f in films if f["slug"] in args.holdout]
+    tag = "TRAIN-FIT (all 9)" if args.train_all else "HELD-OUT"
+    print(f"\n=== {tag} (C++ features, knee, ±{args.tol}s) ===")
+    print(f"{'film':26s} {'TP':>4}{'FP':>5}{'FN':>5} {'P':>5}{'R':>5}{'F1':>5}")
+    f1s = []
+    for f in ev:
+        X, y, yb, ts = load(f["slug"], f["xray"])
+        prob = np.clip(reg.predict(X), 0, 1)
+        peaks = knee_boundaries(prob)
+        Tset = np.where(yb > 0.5)[0]
+        P, R, F, tp, fp, fn = prf(peaks, Tset, args.tol)
+        f1s.append(F)
+        print(f"{f['slug'][:26]:26s} {tp:>4}{fp:>5}{fn:>5} {P*100:4.0f}%{R*100:4.0f}%{F*100:4.0f}%")
+    print(f"\nmacro-F1: {np.mean(f1s)*100:.1f}%")
+    if args.train_all:
+        reg.save_model(str(Path(args.out) / "xgb_boundary_cpp.json"))
+        print(f"[xgb-cpp] shipped model → {args.out}/xgb_boundary_cpp.json", file=sys.stderr)
+
+
+if __name__ == "__main__":
+    main()

scripts/validation/sample_eval.py

@@ -61,7 +61,15 @@ class Prediction:
             keys = keys_for(imdb_id=a.get("imdb_id"), tmdb_id=a.get("tmdb_id"),
                             jellyfin_id=a.get("jellyfin_id"), name=a.get("name"),
                             crosswalk=crosswalk)
-            windows = [(float(t0), float(t1)) for t0, t1 in a.get("scenes", [])]
+            # schema_version 1: scenes is [[t0, t1], ...]  (list of pairs)
+            # schema_version 2: scenes is [{"start":…, "end":…, "belief":…, …}, …]
+            windows = []
+            for s in a.get("scenes", []):
+                if isinstance(s, dict):
+                    windows.append((float(s["start"]), float(s["end"])))
+                else:
+                    t0, t1 = s[0], s[1]
+                    windows.append((float(t0), float(t1)))
             for _, t1 in windows:
                 self._max_t = max(self._max_t, t1)
             self.actors.append({"keys": keys, "windows": windows})

src/backends/ort_provider.hpp

@@ -124,8 +124,21 @@ inline OrtProvider apply_ort_provider(Ort::SessionOptions& opts,
         try {
             OrtROCMProviderOptions rocm{};
             rocm.device_id = 0;
+            // Without these MIOpen runs convolutions on the no-workspace GEMM
+            // fallback (the "GemmFwdRest, provided ptr: 0 size: 0" warnings), which
+            // is the slow path — most visible on the conv-heavy TransNetV2 scene
+            // detector. Exhaustive search lets MIOpen pick the fast conv kernel,
+            // and TunableOp autotunes the GEMMs; both cache to the MIOpen user DB
+            // (MIOPEN_USER_DB_PATH), so the tuning cost is paid once per shape.
+            // Opt-out via SAE_ROCM_NOTUNE=1 for a quick no-warmup run.
+            const bool tune = std::getenv("SAE_ROCM_NOTUNE") == nullptr;
+            rocm.miopen_conv_exhaustive_search = tune ? 1 : 0;
+            rocm.tunable_op_enable             = tune;
+            rocm.tunable_op_tuning_enable      = tune;
             opts.AppendExecutionProvider_ROCM(rocm);
-            std::cerr << "[" << label << "] ROCm provider\n";
+            std::cerr << "[" << label << "] ROCm provider"
+                      << (tune ? " (MIOpen exhaustive + TunableOp)" : " (untuned)")
+                      << "\n";
             return OrtProvider::ROCm;
         } catch (const Ort::Exception& e) {
             std::cerr << "[" << label << "] ROCm unavailable ("

src/config.hpp

@@ -11,6 +11,20 @@ enum class Verbosity {
     standard,  // per-frame detail: bbox, similarity, unknowns logged
     xray,      // Jellyfin-Xray format: {"second": ["Actor", ...], ...}
 };
+
+// How a track's accepted frames become a reported presence window.
+enum class PresenceMode {
+    // A claim IS its track's [first_seen, last_seen] (AR-012/AR-013). The
+    // default and the only mode whose semantics the register validated.
+    track_extent,
+    // Flood-fill: snap each claim to the shot it sits in, so an actor seen once
+    // anywhere in a scene is reported for the whole scene [prev_boundary,
+    // next_boundary]. Trades precision for recall against X-Ray's per-scene cast
+    // granularity. Snaps to TransNetV2 shot boundaries (is_scene_boundary) when a
+    // scene detector populated them, else to the always-on histogram cuts
+    // (is_cut). With no boundaries at all it degrades to track_extent per claim.
+    flood,
+};
 // debug verbosity = compile with -DSAE_DEBUG → scene_analyze_debug binary
 
 struct Config {
@@ -71,40 +85,26 @@ struct Config {
     std::string arcface_model;
     std::string arcface_engine;        // optional path to a pre-built TRT engine; bypasses ORT
     int         embed_batch_size{4};    // max faces per ORT Run() call — bounds per-call latency
-    float       match_prior{0.5f};        // base-rate prior; 0.5 = use calibrated sigmoid directly
+    float       match_prior{0.433f};      // base-rate prior; 10-knob DE optimum (was 0.5)
     // Tuned by Differential Evolution against Amazon X-Ray per-second presence
-    // over the 4-film rep4 matrix. Best model+mode: LVFace-B_Glint360K, full
+    // over ALL 9 films (opencv5 build, LVFace-B_Glint360K, full gallery,
-    // gallery, expansion on. Supersedes an earlier 9-film scene-union tuning
+    // expansion on), a 10-parameter sweep — see docs/model-bakeoff.md. The
-    // (0.76); that metric hid out-of-cast false positives.
+    // per-second misID-weighted macro-F1 optimum is 64.0% (P 79.0%, R 61.1%).
     //
-    // **Read the provenance before trusting the value.** Two things about it:
+    // This is a permissive operating point: the sweep discovered that with
+    // flood-fill presence recovering recall, a LOW threshold pays off. It
+    // supersedes the earlier 0.754, which came from a 4-film subset under the
+    // now-withdrawn anneal/extinction windows and was never re-derived after a
+    // scoring-bug fix. The full-9-film sweep at 0.485 beats it.
     //
-    // 1. The document it came from no longer exists under that name. It was
+    // Caveat, still true: the optimum generalises unevenly. It is strong on 7 of
-    //    docs/rep4-optimizer-results.md, renamed to docs/model-bakeoff.md and
+    // 9 films (F1 62–80%) and weak on two — The Many Saints of Newark (an
-    //    then rewritten (0bd2747). This comment pointed at the dead path for
+    // ensemble of look-alikes; nearly all the run's misIDs land here) and
-    //    long enough that the number looked unsourced. The original is still
+    // Scarface (sparse cuts, so flood-fill over-extends: R 95% / P 26%). Both
-    //    readable at `git show d340da7:docs/rep4-optimizer-results.md`, where
+    // were the low outliers in every prior run too. Shipped because it wins on
-    //    the shipped triple appears as
+    // average and on the misID-weighted objective; not a settled, film-agnostic
-    //    `prob_threshold=0.754, anneal_sec=35.5`.
+    // constant.
-    //
+    float       prob_threshold{0.485f};   // posterior P(match | sim, prior) threshold
-    // 2. **0.754 predates a scoring bug fix and was never re-derived.** That
-    //    same rewrite reports finding "a real scoring bug in optimize.py: a
-    //    candidate whose hardest film's replay timed out was averaged over
-    //    survivors instead of penalized, silently rewarding partial coverage.
-    //    Affected 3 of 16 training combos". The corrected sweep converged
-    //    somewhere else — the surviving document records anneal_sec=59.2,
-    //    extinction_sec=59.2 against the 35.5/57.4 shipped alongside this
-    //    threshold — and no corrected prob_threshold is recorded anywhere.
-    //    (The other two constants are now withdrawn outright, which is why
-    //    only this one still matters.)
-    //
-    // The doc is also candid that the optimum "generalizes unevenly — strong on
-    // 3 of 5 held-out films, badly broken on 2 (one with a 974-count misID
-    // blowup)", and that it is shipped anyway because it still beats the old
-    // defaults on average. That is a defensible call and not a settled,
-    // film-agnostic optimum; it should be visible here rather than only in a
-    // document this comment used to point at incorrectly.
-    float       prob_threshold{0.754f};   // posterior P(match | sim, prior) threshold
     // TRACES: AR-024 | SR-002
     // match_threshold (0.45), match_ratio (0.80) and match_ratio_ceil (0.65) are
     // RETIRED, joining track_max_embed_dist, cut_revive_sim, expand_novelty_sim
@@ -115,6 +115,22 @@ struct Config {
     // fallback everywhere, which is at least the same wrong number in every
     // stage. See identity_matcher_node.hpp.
 
+    // ── Presence derivation ──────────────────────────────────────────────────
+    // How accepted frames become a reported window. flood requires scene_detect.
+    // Default flood: the 10-knob DE optimum uses it — snapping presence to the
+    // shot recovers enough recall against X-Ray's scene-level cast to win the
+    // misID-weighted F1, at a precision cost that is a net gain on 7 of 9 films.
+    // Falls back to track_extent per claim when no boundaries exist. See
+    // docs/model-bakeoff.md and PresenceMode above.
+    PresenceMode presence_mode{PresenceMode::flood};
+
+    // Path to the learned XGBoost scene-boundary model. When set (build has
+    // SAE_SCENE_XGB), the camera-position node stamps a per-frame RGB histogram
+    // and the sink runs the detector post-EOF to supply flood-fill boundaries —
+    // the measured best flood boundary source (presence F1 ~76% vs ~64% for the
+    // always-on histogram cut). Empty → flood falls back to is_cut.
+    std::string  scene_xgb_model;
+
     // ── Cut detection ────────────────────────────────────────────────────────
     float cut_threshold{0.70f};         // grayscale histogram correlation below this → hard cut
 
@@ -162,7 +178,7 @@ struct Config {
     // frame-dependent (AR-007): on is_cut / is_scene_boundary, and for any track
     // that is no longer on screen, it drops to 0 (embedding only), because
     // position carries no information across a viewpoint change or a gap.
-    float track_alpha{0.4f};            // base cost weight: 0=embedding only, 1=spatial only
+    float track_alpha{0.435f};          // base cost weight: 0=embedding only, 1=spatial only (10-knob DE optimum)
     float track_min_iou{0.1f};          // IoU below which spatial link alone is rejected
     // Minimum P(same person) for an association to be admissible on appearance
     // alone. This replaces track_max_embed_dist (a raw cosine distance, AR-024).
@@ -175,7 +191,7 @@ struct Config {
     // Replaces track_max_frames_missing: a frame count silently changed meaning
     // with sample_fps, and the same number had to be guessed twice (once for an
     // ordinary miss, once for a cut). Seconds mean one thing at any sample rate.
-    double track_extinction_sec{5.0};
+    double track_extinction_sec{31.0};   // 10-knob DE optimum (was 5.0)
 
     // ── Ownership and evidence accumulation (AR-025) ──────────────────────────
     // TRACES: AR-025, AR-017 | SR-002
@@ -189,8 +205,10 @@ struct Config {
     // ownership_logodds is arguably the most consequential constant in the
     // pipeline after prob_threshold: below it a track produces no presence
     // claim at all, so it decides whether an actor is reported rather than how
-    // confidently. 2.0 is a posterior of ~0.88. Unswept.
+    // confidently. 1.72 is a posterior of ~0.85 — the 10-knob DE optimum (was
-    float  ownership_logodds{2.0f};
+    // an unswept 2.0 ≈ 0.88); slightly more permissive, consistent with the
+    // low-threshold operating point the sweep converged on.
+    float  ownership_logodds{1.72f};
 
     // How much a single observation may move a track's belief. n_eff =
     // n / (1 + (n-1)·rho), so rho_max caps what a repeated view can ever be
@@ -199,10 +217,11 @@ struct Config {
     // detection, alignment and noise realisation, so a little independent
     // evidence survives. Setting it to 1 freezes belief after the first frame,
     // which is the bug this replaced.
-    float  evidence_rho_max{0.5f};
+    float  evidence_rho_max{0.204f};   // 10-knob DE optimum (was 0.5): weights a
+                                       // held pose closer to a single observation
     // P(same view) below this and the observation counts as a genuinely new
     // look, so it joins the per-track view set.
-    float  evidence_admit_below{0.6f};
+    float  evidence_admit_below{0.784f};  // 10-knob DE optimum (was 0.6)
     // Distinct views remembered per track, which bounds the novelty comparison.
     int    evidence_max_views{8};
 
@@ -250,10 +269,10 @@ struct Config {
     // at promotion time — see track_gallery.hpp. This is the only threshold the
     // expansion path has: it replaces the raw-cosine expand_novelty_sim (0.55)
     // and expand_track_spread_max (0.60), which are retired (AR-024).
-    // Working values pending VR-007; sweep both bounds, they fail in opposite
+    // 10-knob DE optimum (was 0.90/0.95). The sweep widened the band — a lower lo
-    // directions.
+    // admits more pose-varied views into the annex — which the optimum preferred.
-    float  expand_band_lo{0.90f};
+    float  expand_band_lo{0.804f};
-    float  expand_band_hi{0.95f};
+    float  expand_band_hi{0.952f};
     int    expand_min_anchor_frames{3};  // require ≥N accepted frames naming the actor before
                                          //   the track is confirmed and its buffer promoted
     std::string expand_debug_dir;        // if set, dump promoted mugshots + embeddings here

src/dump_embeddings.cpp

@@ -8,6 +8,12 @@
 // gallery file needed. Purpose-built for the optimizer's replay corpus and the
 // embedding-model bake-off (dump each --arcface model over the film set).
 //
+// The camera-position (histogram) cut detector runs, so frames/is_cut is recorded
+// in the dump. Flood-fill presence snaps to those cuts. TransNetV2 scene detection
+// is NOT run here: on the ROCm build it needs MIGraphX, whose statically-linked
+// LLVM collides with the VAAPI decoder's system libLLVM and crashes at compile, so
+// the two cannot share a process. is_scene_boundary therefore stays 0 here.
+//
 // Usage:
 //   dump_embeddings --movie <path> --out <dump.h5> [--arcface <model.onnx>]
 //                   [--detector <model.onnx>] [--fps 1] [--start S] [--end S]

src/inference/audio_logpsd.hpp

@@ -0,0 +1,143 @@
+#pragma once
+// Per-second audio log-PSD, C++ parity with scripts/scene_detector/
+// extract_audio_features.py — the audio tower input for the XGBoost scene
+// detector. Decodes the whole track to mono 16 kHz, then one FFT per second over
+// a 4 s Hann-windowed window, power pooled into geomspace log-frequency bands,
+// L1-normalised (shape not loudness) and log1p-compressed.
+//
+// Must match the Python exactly (SR=16000, WIN_SEC=4, N_BINS=64→geomspace unique
+// edges, log1p(band*1e3)); the shipped model was trained on those features.
+
+extern "C" {
+#include <libavformat/avformat.h>
+#include <libavcodec/avcodec.h>
+#include <libavutil/opt.h>
+#include <libswresample/swresample.h>
+}
+#include <fftw3.h>
+
+#include <algorithm>
+#include <cmath>
+#include <cstdint>
+#include <stdexcept>
+#include <string>
+#include <vector>
+
+class AudioLogPSD {
+public:
+    static constexpr int   kSR = 16000;
+    static constexpr double kHop = 1.0;    // 1 feature row / second
+    static constexpr double kWin = 4.0;    // FFT window seconds
+    static constexpr int   kNBins = 64;    // geomspace target (dedups to ~57)
+
+    // Returns [T][B] per-second log-PSD (T ≈ film seconds, B ≈ 57), aligned to the
+    // 1 fps grid. Empty on decode failure (caller then feeds a zero block).
+    static std::vector<std::vector<float>> extract(const std::string& path) {
+        std::vector<float> mono = decode_mono_16k(path);
+        if (mono.empty()) return {};
+        return features(mono);
+    }
+
+    // Public for the parity harness.
+    static std::vector<std::vector<float>> features(const std::vector<float>& mono) {
+        const int win = int(kSR * kWin), hop = int(kSR * kHop);
+        const int T = int(mono.size()) / hop;
+        if (T <= 0) return {};
+        const int nfreq = win/2 + 1;
+        std::vector<int> edges = geomspace_edges(nfreq);
+        const int nb = int(edges.size()) - 1;
+
+        // Hann window (matches scipy.signal.windows.hann, sym=True default → but
+        // numpy code uses sps.windows.hann(win) which is symmetric).
+        std::vector<double> hann(win);
+        for (int i = 0; i < win; ++i)
+            hann[i] = 0.5 - 0.5*std::cos(2.0*M_PI*i/(win-1));
+
+        std::vector<double> in(win);
+        auto* out = fftw_alloc_complex(nfreq);
+        fftw_plan plan = fftw_plan_dft_r2c_1d(win, in.data(), out, FFTW_ESTIMATE);
+
+        std::vector<std::vector<float>> feat(T, std::vector<float>(nb, 0.f));
+        const int half = win/2;
+        for (int t = 0; t < T; ++t) {
+            int centre = t*hop + hop/2;
+            int s = centre - half;
+            for (int i = 0; i < win; ++i) {
+                int idx = s + i;
+                double v = (idx >= 0 && idx < int(mono.size())) ? mono[idx] : 0.0;
+                in[i] = v * hann[i];
+            }
+            fftw_execute(plan);
+            // power spectrum + 1e-12
+            std::vector<double> psd(nfreq);
+            for (int i = 0; i < nfreq; ++i)
+                psd[i] = out[i][0]*out[i][0] + out[i][1]*out[i][1] + 1e-12;
+            std::vector<double> band(nb, 0.0);
+            double tot = 0.0;
+            for (int b = 0; b < nb; ++b) {
+                for (int i = edges[b]; i < edges[b+1]; ++i) band[b] += psd[i];
+                tot += band[b];
+            }
+            for (int b = 0; b < nb; ++b)
+                feat[t][b] = float(std::log1p(band[b]/tot * 1e3));
+        }
+        fftw_destroy_plan(plan); fftw_free(out);
+        return feat;
+    }
+
+private:
+    // np.unique(np.geomspace(1, nfreq-1, N_BINS+1).astype(int))
+    static std::vector<int> geomspace_edges(int nfreq) {
+        const int n = kNBins + 1;
+        double a = std::log(1.0), b = std::log(double(nfreq-1));
+        std::vector<int> raw(n);
+        for (int i = 0; i < n; ++i)
+            raw[i] = int(std::exp(a + (b-a)*i/(n-1)));   // .astype(int) truncates
+        std::vector<int> uniq;
+        for (int v : raw) if (uniq.empty() || v != uniq.back()) uniq.push_back(v);
+        return uniq;
+    }
+
+    static std::vector<float> decode_mono_16k(const std::string& path) {
+        AVFormatContext* fmt = nullptr;
+        if (avformat_open_input(&fmt, path.c_str(), nullptr, nullptr) < 0) return {};
+        std::vector<float> out;
+        SwrContext* swr = nullptr; AVCodecContext* dec = nullptr;
+        AVPacket* pkt = av_packet_alloc(); AVFrame* fr = av_frame_alloc();
+        try {
+            if (avformat_find_stream_info(fmt, nullptr) < 0) throw 0;
+            int ai = av_find_best_stream(fmt, AVMEDIA_TYPE_AUDIO, -1, -1, nullptr, 0);
+            if (ai < 0) throw 0;
+            AVStream* st = fmt->streams[ai];
+            const AVCodec* codec = avcodec_find_decoder(st->codecpar->codec_id);
+            dec = avcodec_alloc_context3(codec);
+            avcodec_parameters_to_context(dec, st->codecpar);
+            if (avcodec_open2(dec, codec, nullptr) < 0) throw 0;
+
+            AVChannelLayout out_ch = AV_CHANNEL_LAYOUT_MONO;
+            swr_alloc_set_opts2(&swr, &out_ch, AV_SAMPLE_FMT_FLT, kSR,
+                                &dec->ch_layout, dec->sample_fmt,
+                                dec->sample_rate ? dec->sample_rate : kSR, 0, nullptr);
+            if (!swr || swr_init(swr) < 0) throw 0;
+
+            while (av_read_frame(fmt, pkt) >= 0) {
+                if (pkt->stream_index == ai && avcodec_send_packet(dec, pkt) >= 0) {
+                    while (avcodec_receive_frame(dec, fr) >= 0) {
+                        int max_out = swr_get_out_samples(swr, fr->nb_samples);
+                        size_t base = out.size(); out.resize(base + max_out);
+                        uint8_t* dst = reinterpret_cast<uint8_t*>(out.data() + base);
+                        int got = swr_convert(swr, &dst, max_out,
+                            (const uint8_t**)fr->extended_data, fr->nb_samples);
+                        out.resize(base + std::max(0, got));
+                    }
+                }
+                av_packet_unref(pkt);
+            }
+        } catch (...) { out.clear(); }
+        if (swr) swr_free(&swr);
+        if (dec) avcodec_free_context(&dec);
+        av_frame_free(&fr); av_packet_free(&pkt);
+        avformat_close_input(&fmt);
+        return out;
+    }
+};

src/inference/xgb_scene_boundary.hpp

@@ -0,0 +1,319 @@
+#pragma once
+// XGBoost scene-boundary detector — C++ inference of the shipped model
+// (models/scene_boundary_xgb.json), for flood-fill presence in the live pipeline.
+//
+// This is a POST-EOF step (like flood-fill itself): the per-film knee threshold
+// needs every peak, so boundaries can only be finalized after the whole film is
+// seen. The result sink collects a per-frame RGB histogram; at EOF it calls
+// boundaries() with the full (timestamp, hist) series and gets back the boundary
+// timestamps to flood-snap against.
+//
+// The feature pipeline MUST match scripts/scene_detector/train_scene_boundary.py
+// exactly (206 features): a ±WIN=3s window of per-second base features + a
+// 3-value debounce clock. Base per second (29):
+//   video(17): sym-delta |hist(t+k)-hist(t-k)| L1 at k=1,2,4,8; per-channel corr
+//              to t-1 (3); ramp bank at H=2,4,6,8,10 on z-normed hist (5);
+//              per-channel energy (3); debounce phase/decay from |delta k=1| (2)
+//   audio(12): same but on the log-PSD, no corr, 1 energy   [ZERO when no audio]
+// then window flatten t-3..t+3 (×7) and append clock (dt, phase, decay).
+//
+// Audio is not available live (the pipeline has no per-second PSD stream), so the
+// audio block is fed zeros — the model was trained with audio present but it is
+// weak (measured) and XGBoost tolerates a constant block; the video signal
+// carries the detector. (If live audio is added later, fill the block.)
+
+#include <xgboost/c_api.h>
+
+#include <algorithm>
+#include <array>
+#include <cmath>
+#include <cstdint>
+#include <numeric>
+#include <stdexcept>
+#include <string>
+#include <vector>
+
+class XGBSceneBoundary {
+public:
+    // Must match kHistBins in embedding_dump_node.hpp / the training dump.
+    static constexpr int kHistBins = 32;          // per channel → 96-float hist
+    static constexpr int kWin = 3;                // ±WIN-second window
+    static constexpr double kSigmaTau = 205.0;    // SCENE_TAU (unused at infer; kept for parity docs)
+    static constexpr int    kRampScales[5] = {2, 4, 6, 8, 10};
+
+    explicit XGBSceneBoundary(const std::string& model_path) {
+        if (XGBoosterCreate(nullptr, 0, &booster_) != 0)
+            throw std::runtime_error("XGBoosterCreate failed");
+        if (XGBoosterLoadModel(booster_, model_path.c_str()) != 0)
+            throw std::runtime_error("XGBoosterLoadModel failed: " +
+                                     std::string(XGBGetLastError()));
+    }
+    ~XGBSceneBoundary() { if (booster_) XGBoosterFree(booster_); }
+    XGBSceneBoundary(const XGBSceneBoundary&) = delete;
+    XGBSceneBoundary& operator=(const XGBSceneBoundary&) = delete;
+
+    // hist:  T rows × 96 (normalised RGB histogram per second).
+    // audio: T rows × B log-PSD (from AudioLogPSD; aligned to the same seconds),
+    //        or empty → the audio block is filled with its zero-input values
+    //        (deltas/ramp/energy 0, but debounce phase=1/decay=exp(-1), matching
+    //        the Python audio_features on a zero series).
+    // Returns boundary timestamps (knee-selected).
+    std::vector<double> boundaries(const std::vector<std::vector<float>>& hist,
+                                   const std::vector<double>& ts,
+                                   const std::vector<std::vector<float>>& audio = {}) {
+        const int T = static_cast<int>(hist.size());
+        if (T < 2 * kWin + 2) return {};
+        auto base = build_base(hist, audio);           // [T][29]
+        std::vector<float> X = window_and_clock(base, hist);
+        std::vector<float> prob = predict(X, T, 206);
+        return knee_boundaries(prob, ts);
+    }
+
+    static std::vector<std::vector<float>> debug_base(const std::vector<std::vector<float>>& hist,
+                                                      const std::vector<std::vector<float>>& audio = {}) {
+        return build_base(hist, audio);
+    }
+
+    // Predict boundaries from a precomputed [rows×cols] feature matrix (for the
+    // clean parity check: same bytes both sides).
+    std::vector<double> boundaries_from_features(const std::vector<float>& X, int rows,
+                                                 int cols, const std::vector<double>& ts) {
+        auto prob = predict(X, rows, cols);
+        return knee_boundaries(prob, ts);
+    }
+    std::vector<float> debug_predict(const std::vector<float>& X, int r, int c) {
+        return predict(X, r, c);
+    }
+    static std::vector<int> debug_find_peaks(const std::vector<float>& p, int d) {
+        return find_peaks(p, d);
+    }
+
+    // The flat [T*206] feature matrix — exposed so TRAINING uses the exact same
+    // C++ features as inference (parity by construction; no numpy re-match). The
+    // Python trainer reshapes to [T,206], attaches the soft target, and fits.
+    static std::vector<float> feature_matrix(const std::vector<std::vector<float>>& hist,
+                                             const std::vector<std::vector<float>>& audio) {
+        auto base = build_base(hist, audio);
+        return window_and_clock(base, hist);
+    }
+    static constexpr int kNFeatures = 206;
+
+private:
+    BoosterHandle booster_{nullptr};
+
+    // ── feature builders (exact parity with the Python) ──────────────────────
+
+    static float l1(const std::vector<float>& a, const std::vector<float>& b) {
+        float s = 0; for (size_t i = 0; i < a.size(); ++i) s += std::fabs(a[i] - b[i]);
+        return s;
+    }
+
+    // z-normalise each of the 96 columns across time (matches _znorm).
+    static std::vector<std::vector<float>> znorm(const std::vector<std::vector<float>>& h) {
+        const int T = h.size(), D = h[0].size();
+        std::vector<float> mu(D, 0), sd(D, 0);
+        for (auto& r : h) for (int d = 0; d < D; ++d) mu[d] += r[d];
+        for (int d = 0; d < D; ++d) mu[d] /= T;
+        for (auto& r : h) for (int d = 0; d < D; ++d) sd[d] += (r[d]-mu[d])*(r[d]-mu[d]);
+        for (int d = 0; d < D; ++d) sd[d] = std::sqrt(sd[d]/T) + 1e-6f;
+        std::vector<std::vector<float>> z(T, std::vector<float>(D));
+        for (int t = 0; t < T; ++t) for (int d = 0; d < D; ++d) z[t][d] = (h[t][d]-mu[d])/sd[d];
+        return z;
+    }
+
+    // ramp bank: L2 of the antisymmetric ramp-weighted sum over ±H, per scale.
+    // Matches ramp_bank() (np.convolve 'same' with reversed kernel; sign folds
+    // into the L2 norm so the direct antisymmetric sum is equivalent).
+    static std::vector<std::array<float,5>> ramp_bank(const std::vector<std::vector<float>>& z) {
+        const int T = z.size(), D = z[0].size();
+        std::vector<std::array<float,5>> out(T);
+        for (int k = 0; k < 5; ++k) {
+            const int H = kRampScales[k];
+            for (int t = 0; t < T; ++t) {
+                std::vector<double> acc(D, 0.0);
+                for (int l = -H; l <= H; ++l) {
+                    int idx = t + l;
+                    if (idx < 0 || idx >= T) continue;
+                    double w = (l == 0) ? 0.0 : (l > 0 ? 1.0 : -1.0) * (double(std::abs(l))/H);
+                    for (int d = 0; d < D; ++d) acc[d] += w * z[idx][d];
+                }
+                double n = 0; for (double v : acc) n += v*v;
+                out[t][k] = static_cast<float>(std::sqrt(n));
+            }
+        }
+        return out;
+    }
+
+    // Generic symmetric-delta + ramp + energy + debounce feature block for one
+    // modality's z-normable series `raw` (hist or PSD). Fills `out` columns
+    // [off .. off+width). corr=true adds the 3 per-channel corr features (video
+    // only); n_energy is 3 (video, per-channel) or 1 (audio, total).
+    static void modality_block(const std::vector<std::vector<float>>& raw,
+                               bool corr, int n_energy,
+                               std::vector<std::vector<float>>& out, int off) {
+        const int T = raw.size();
+        auto z = znorm(raw);
+        auto rb = ramp_bank(z);
+        auto sym = [&](int t, int k)->float{
+            int f = std::min(T-1, t+k), b = std::max(0, t-k);
+            return l1(raw[f], raw[b]);
+        };
+        const int B = kHistBins;   // only used for corr (video)
+        for (int t = 0; t < T; ++t) {
+            int o = off;
+            for (int k : {1,2,4,8}) out[t][o++] = sym(t,k);
+            if (corr) {
+                int tp = std::max(0, t-1);
+                for (int c = 0; c < 3; ++c) {
+                    double ma=0, mb=0;
+                    for (int i=0;i<B;++i){ ma+=raw[t][c*B+i]; mb+=raw[tp][c*B+i]; }
+                    ma/=B; mb/=B; double num=0, da=0, db=0;
+                    for (int i=0;i<B;++i){ double x=raw[t][c*B+i]-ma, y=raw[tp][c*B+i]-mb;
+                        num+=x*y; da+=x*x; db+=y*y; }
+                    out[t][o++] = float(num/(std::sqrt(da*db)+1e-9));
+                }
+            }
+            for (int k=0;k<5;++k) out[t][o++] = rb[t][k];
+            if (n_energy == 3) {
+                for (int c=0;c<3;++c){ float e=0; for(int i=0;i<B;++i) e+=raw[t][c*B+i]; out[t][o++]=e; }
+            } else {
+                float e=0; for (float v : raw[t]) e+=v; out[t][o++]=e;
+            }
+            o += 2;   // debounce filled below
+        }
+        // debounce from this block's delta-k1 (its first column = off)
+        std::vector<float> d1(T); for (int t=0;t<T;++t) d1[t]=out[t][off];
+        auto clk = debounce_phase(d1);
+        // debounce sits at the end of the block: off + 4(deltas) + (corr?3:0) + 5(ramp) + n_energy
+        int deb = off + 4 + (corr?3:0) + 5 + n_energy;
+        for (int t=0;t<T;++t){ out[t][deb]=clk[t].first; out[t][deb+1]=clk[t].second; }
+    }
+
+    // per-second base = video(17) + audio(12). Audio empty → its block is the
+    // zero-series result (deltas/ramp/energy 0, debounce phase=1/decay=exp(-1)).
+    static std::vector<std::vector<float>> build_base(const std::vector<std::vector<float>>& hist,
+                                                      const std::vector<std::vector<float>>& audio) {
+        const int T = hist.size();
+        std::vector<std::vector<float>> base(T, std::vector<float>(29, 0.0f));
+        modality_block(hist, /*corr=*/true, /*n_energy=*/3, base, /*off=*/0);   // video → 0..16
+        if (!audio.empty() && int(audio.size()) == T) {
+            modality_block(audio, /*corr=*/false, /*n_energy=*/1, base, /*off=*/17); // audio → 17..28
+        } else {
+            // zero-series audio: deltas/ramp/energy already 0; only debounce differs.
+            auto clk = debounce_phase(std::vector<float>(T, 0.0f));
+            for (int t=0;t<T;++t){ base[t][27]=clk[t].first; base[t][28]=clk[t].second; }
+        }
+        return base;
+    }
+
+    // matches debounce_phase(): 90th-pct peaks, dt=time since last, phase/decay.
+    static std::vector<std::pair<float,float>> debounce_phase(const std::vector<float>& sig) {
+        const int T = sig.size();
+        std::vector<float> s(sig); std::sort(s.begin(), s.end());
+        float thr = s[std::min(T-1, int(0.90*T))];
+        std::vector<std::pair<float,float>> out(T);
+        int last = -1000000000;
+        for (int t=0;t<T;++t){
+            if (sig[t] > thr) last = t;
+            double dt = (last < -100000000) ? kSigmaTau : double(t - last);
+            out[t] = { float(std::min(1.0, dt/kSigmaTau)), float(std::exp(-dt/kSigmaTau)) };
+        }
+        return out;
+    }
+
+    // window flatten (t-3..t+3, edge-pad) + append the 3-value film clock.
+    static std::vector<float> window_and_clock(const std::vector<std::vector<float>>& base,
+                                               const std::vector<std::vector<float>>& hist) {
+        const int T = base.size(), d = base[0].size();   // d=29
+        // film-level clock: time-since-last-peak on the |delta k1| video signal
+        // (base col 0), same as per_second_matrix's `clock`.
+        std::vector<float> sig(T); for (int t=0;t<T;++t) sig[t]=base[t][0];
+        std::vector<float> ss(sig); std::sort(ss.begin(), ss.end());
+        float thr = ss[std::min(T-1, int(0.90*T))];
+        std::vector<float> X; X.reserve(size_t(T)*206);
+        int last=-1000000000;
+        for (int t=0;t<T;++t){
+            for (int off=-kWin; off<=kWin; ++off){
+                int idx = std::min(T-1, std::max(0, t+off));
+                for (int j=0;j<d;++j) X.push_back(base[idx][j]);
+            }
+            if (sig[t] > thr) last=t;
+            double dt=(last<-100000000)?kSigmaTau:double(t-last);
+            X.push_back(float(dt));
+            X.push_back(float(std::min(1.0, dt/kSigmaTau)));
+            X.push_back(float(std::exp(-dt/kSigmaTau)));
+        }
+        return X;
+    }
+
+    std::vector<float> predict(const std::vector<float>& X, int rows, int cols) {
+        DMatrixHandle dm;
+        if (XGDMatrixCreateFromMat(X.data(), rows, cols, std::nanf(""), &dm) != 0)
+            throw std::runtime_error("XGDMatrixCreateFromMat failed");
+        bst_ulong out_len = 0; const float* out = nullptr;
+        if (XGBoosterPredict(booster_, dm, 0, 0, 0, &out_len, &out) != 0)
+            throw std::runtime_error("XGBoosterPredict failed");
+        std::vector<float> p(out, out + out_len);
+        XGDMatrixFree(dm);
+        for (auto& v : p) v = std::clamp(v, 0.f, 1.f);
+        return p;
+    }
+
+    // Exact replica of scipy.signal.find_peaks(x, distance=d):
+    //   1. local maxima (plateau-aware: rising then falling, midpoint of a flat top)
+    //   2. keep peaks by DESCENDING height; drop any within `d` of an already-kept
+    //      taller peak. This is height-priority, NOT the greedy left-to-right merge
+    //      — the two give different peak sets and hence a different knee.
+    static std::vector<int> find_peaks(const std::vector<float>& x, int d) {
+        const int n = x.size();
+        std::vector<int> mid;
+        int i = 1;
+        while (i < n-1) {
+            if (x[i-1] < x[i]) {
+                int ahead = i+1;
+                while (ahead < n-1 && x[ahead] == x[i]) ahead++;
+                if (x[ahead] < x[i]) mid.push_back((i + ahead - 1) / 2);
+                i = ahead;
+            } else i++;
+        }
+        // height-priority distance filter (scipy's _select_by_peak_distance)
+        std::vector<int> order(mid.size());
+        for (size_t k = 0; k < mid.size(); ++k) order[k] = k;
+        std::sort(order.begin(), order.end(),
+                  [&](int a, int b){ return x[mid[a]] < x[mid[b]]; });  // ascending
+        std::vector<char> keep(mid.size(), 1);
+        for (int j = int(order.size())-1; j >= 0; --j) {   // tallest first
+            int k = order[j];
+            if (!keep[k]) continue;
+            for (int l = k-1; l >= 0 && mid[k]-mid[l] < d; --l) keep[l] = 0;
+            for (int r = k+1; r < int(mid.size()) && mid[r]-mid[k] < d; ++r) keep[r] = 0;
+        }
+        std::vector<int> out;
+        for (size_t k = 0; k < mid.size(); ++k) if (keep[k]) out.push_back(mid[k]);
+        return out;
+    }
+
+    // knee threshold on peak heights → boundary timestamps (matches knee_boundaries).
+    static std::vector<double> knee_boundaries(const std::vector<float>& prob,
+                                               const std::vector<double>& ts,
+                                               int min_gap = 5) {
+        std::vector<int> pk = find_peaks(prob, min_gap);
+        if (pk.size() < 5) {
+            std::vector<double> r; for (int i : pk) r.push_back(ts[i]); return r;
+        }
+        std::vector<float> h; for (int i : pk) h.push_back(prob[i]);
+        std::sort(h.begin(), h.end(), std::greater<float>());
+        int n = h.size(); float h0 = h.front() + 1e-9f;
+        int kbest = 0; double dmax = -1;
+        for (int i = 0; i < n; ++i) {
+            double x = double(i)/(n-1);
+            double yv = h[i]/h0;
+            double chord = (h[0]/h0) + ((h[n-1]/h0)-(h[0]/h0))*x;
+            if (chord - yv > dmax) { dmax = chord - yv; kbest = i; }
+        }
+        float knee = h[kbest];
+        std::vector<double> out;
+        for (int i : pk) if (prob[i] >= knee) out.push_back(ts[i]);
+        return out;
+    }
+};

src/kpn_bindings.cpp

@@ -1,11 +1,38 @@
-// sae_kpn — run the real downstream pipeline nodes (face_tracker, identity_matcher,
+// sae_kpn — run the real downstream pipeline inside a Python-assembled KPN
-// frame_annotation) inside a Python-assembled KPN network, fed by a Python HDF5 replay
+// network, fed by a Python HDF5 replay source. Lets a parameter sweep re-run the
-// source. Lets a parameter sweep re-run the exact C++ matching/tracking logic over
+// exact C++ tracking/matching/presence logic over dumped embeddings — no video
-// dumped embeddings — no video decode, no GPU — with different Config knobs each run.
+// decode, no GPU — with different Config knobs each run.
+//
+/// TRACES: VR-011, VR-002 | PR-002
+//
+// **The whole chain is C++, including the sink.** That is the VR-011 change and
+// it is the point of the requirement: replay must drive the real nodes, not a
+// reimplementation. Two things were wrong before.
+//
+//   1. It did not compile. `add_face_tracker` built `FaceTrackerFunc` from a
+//      Config alone, and the tracker has required a TrackRegistry and a
+//      calibration since AR-007/AR-008 moved association into probability
+//      space. Any .so in a stale build/ predates that.
+//
+//   2. Presence was rebuilt in Python. `replay.py::build_minimal` merged
+//      per-frame detections into windows by annealing gaps — which is what the
+//      pipeline did before AR-012. The sink now builds a window from a
+//      TrackRegistry claim: the extent of a track an actor owned, starting when
+//      they appeared rather than when recognition first succeeded. Those answer
+//      different questions, so every sweep was tuning against a contract the
+//      shipped code had stopped honouring.
+//
+// Both had the same root cause, which is why this is one binding and not three.
+// The chain has a construction ORDER — the matcher fits the calibration, the
+// registry needs a discounter built from it, the tracker needs both, and the
+// sink needs the registry's claims — and a factory-per-node API cannot express
+// it. `add_pipeline` mirrors main.cpp exactly and is the only way to build the
+// chain, so the ordering cannot be got wrong again from Python.
 //
 // Boundary types (cross the Python seam):
 //   EmbeddedSceneFrame  IN   (built by the Python replay source from HDF5 arrays)
-//   SceneAnnotation     OUT  (read by the Python sink → presence JSON)
+//   SceneAnnotation     OUT  (optional tee for per-frame debug rendering only —
+//                             the presence output is written by the C++ sink)
 // Intermediate types (TrackedSceneFrame, MatchedSceneFrame) flow C++→C++ only, but
 // still need channel factories + converters registered so PyNetwork can wire them.
 
@@ -20,6 +47,9 @@
 #include "nodes/face_tracker_node.hpp"
 #include "nodes/identity_matcher_node.hpp"
 #include "nodes/frame_annotation_node.hpp"
+#include "nodes/result_sink_node.hpp"
+#include "track_registry.hpp"
+#include "evidence_discount.hpp"
 
 #include <nanobind/nanobind.h>
 #include <nanobind/ndarray.h>
@@ -27,6 +57,8 @@
 #include <nanobind/stl/vector.h>
 #include <nanobind/stl/map.h>
 
+#include <atomic>
+#include <map>
 #include <memory>
 #include <optional>
 #include <variant>
@@ -34,10 +66,52 @@
 namespace nb = nanobind;
 using namespace nb::literals;
 
+// ── ReplaySession ─────────────────────────────────────────────────────────────
+/// TRACES: VR-011 | PR-002
+/// State the network's nodes reference but do not own.
+///
+/// ResultSinkFunc holds `std::atomic<bool>&`, exactly as it does under main(),
+/// where it is a stack local in a function that outlives the pipeline. There is
+/// no such frame here -- the network is built and torn down from Python -- so
+/// the flag lives in a session held for the network's lifetime and released
+/// explicitly. The registry is here for the same reason: the sink's claim
+/// callback captures it.
+struct ReplaySession {
+    /// Owns the Config, and must. ResultSinkFunc holds `const Config&` -- under
+    /// main() that is a stack local in a frame which outlives the pipeline, so
+    /// the reference is fine there. There is no such frame here: the network is
+    /// built inside a binding call and torn down from Python, so a Config local
+    /// to add_pipeline dies the moment it returns and the sink is left reading
+    /// freed memory. It presented as an empty output_path -- the sink announced
+    /// `[result_sink] writing ` and wrote nothing.
+    Config                         cfg;
+    std::atomic<bool>              done{false};
+    std::shared_ptr<TrackRegistry> registry;
+};
+
+// Function-local static so ordering against other translation units cannot bite.
+inline std::map<void*, std::shared_ptr<ReplaySession>>& sessions() {
+    static std::map<void*, std::shared_ptr<ReplaySession>> s;
+    return s;
+}
+
 // The variant spanning every type that flows on a channel in the replay chain.
 using SaeVariant = std::variant<EmbeddedSceneFrame, TrackedSceneFrame,
                                 MatchedSceneFrame, SceneAnnotation>;
 
+// ── Node wrapper aliases ──────────────────────────────────────────────────────
+// Named once so add_pipeline and the runtime setters cannot disagree about a
+// node's port names: a mismatch there is a dynamic_cast that returns null, i.e.
+// a runtime setter that silently does nothing.
+using MatcherWrap = kpn::ObjectVariantNodeWrapper<
+    IdentityMatcherFunc, SaeVariant, kpn::in<"tracked">,  kpn::out<"matched">>;
+using TrackerWrap = kpn::ObjectVariantNodeWrapper<
+    FaceTrackerFunc,     SaeVariant, kpn::in<"embedded">, kpn::out<"tracked">>;
+using AnnotWrap   = kpn::ObjectVariantNodeWrapper<
+    FrameAnnotationFunc, SaeVariant, kpn::in<"matched">,  kpn::out<"annotation">>;
+using SinkWrap    = kpn::ObjectVariantNodeWrapper<
+    ResultSinkFunc,      SaeVariant, kpn::in<"annotation">, kpn::out<>>;
+
 // ── Converters ─────────────────────────────────────────────────────────────────
 // Only EmbeddedSceneFrame (in) and SceneAnnotation (out) actually cross the seam;
 // the two intermediates get identity-ish stubs (never converted in practice) so the
@@ -62,6 +136,8 @@ template<> struct PythonConverter<EmbeddedSceneFrame> {
         ef.source.frame_idx     = d.contains("frame_idx") ? nb::cast<int64_t>(d["frame_idx"]) : -1;
         ef.source.eof           = d.contains("eof") ? nb::cast<bool>(d["eof"]) : false;
         ef.source.is_cut        = d.contains("is_cut") ? nb::cast<bool>(d["is_cut"]) : false;
+        ef.source.is_scene_boundary = d.contains("is_scene_boundary")
+                                    ? nb::cast<bool>(d["is_scene_boundary"]) : false;
         if (ef.source.eof) return ef;
 
         // faces: (N,4) bbox, (N,10) landmarks, (N,) confidence, (N,512) embeddings
@@ -180,9 +256,48 @@ static Config config_from_dict(nb::dict d) {
     geti("evidence_max_views",    cfg.evidence_max_views);
     // gallery expansion (usually off for sweeps; expose so it can be toggled)
     if (d.contains("expand_gallery")) cfg.expand_gallery = nb::cast<bool>(d["expand_gallery"]);
+    // AR-018: banded admission bounds for the per-film annex, in probability
+    // space. Reachable from a sweep — the config comment asks for both to be
+    // swept, and they are ignored unless expand_gallery is on. See track_gallery.hpp.
+    getf("expand_band_lo",        cfg.expand_band_lo);
+    getf("expand_band_hi",        cfg.expand_band_hi);
+    // Presence derivation. Accepts a string ("flood"/"track_extent") or a
+    // number (DE only produces floats: >=0.5 → flood) so the sweep can toggle
+    // it as a sixth knob. flood snaps to boundaries in the replayed frames
+    // (is_scene_boundary if present, else is_cut).
+    if (d.contains("presence_mode")) {
+        const auto& pm = d["presence_mode"];
+        bool flood = false;
+        if (nb::isinstance<nb::str>(pm)) flood = (nb::cast<std::string>(pm) == "flood");
+        else                             flood = (nb::cast<double>(pm) >= 0.5);
+        cfg.presence_mode = flood ? PresenceMode::flood : PresenceMode::track_extent;
+    }
+
     /// TRACES: GR-004 | SR-001
     if (d.contains("require_gallery_stamp"))
         cfg.require_gallery_stamp = nb::cast<bool>(d["require_gallery_stamp"]);
+
+    /// TRACES: VR-011 | IR-001 | PR-002 | SR-003
+    // The sink is a real node in this network now, so it needs the two things
+    // that decide what it writes and where. Both used to be irrelevant here
+    // because the replay never had a sink -- Python rebuilt presence instead,
+    // which is the reimplementation VR-002 forbids and VR-011 removes.
+    if (d.contains("output_path"))
+        cfg.output_path = nb::cast<std::string>(d["output_path"]);
+    if (d.contains("verbosity")) {
+        const int v = nb::cast<int>(d["verbosity"]);
+        cfg.verbosity = v == 2 ? Verbosity::xray
+                      : v == 1 ? Verbosity::standard
+                               : Verbosity::minimal;
+    }
+    // Reported verbatim in the truth file's extraction block, so a replayed
+    // manifest says which gallery scope produced it (IR-002).
+    if (d.contains("gallery_scope"))
+        cfg.gallery_scope = nb::cast<std::string>(d["gallery_scope"]);
+    if (d.contains("sample_fps"))
+        cfg.sample_fps = nb::cast<float>(d["sample_fps"]);
+    if (d.contains("movie_path"))
+        cfg.movie_path = nb::cast<std::string>(d["movie_path"]);
     return cfg;
 }
 
@@ -222,38 +337,51 @@ NB_MODULE(sae_kpn, m) {
                             std::move(outs), cap);
     }, "net"_a, "name"_a, "callable"_a, "inputs"_a, "outputs"_a, "capacity"_a = 5);
 
-    // ── Real node factories ─────────────────────────────────────────────────────
+    // ── The pipeline ────────────────────────────────────────────────────────────
-    m.def("add_face_tracker", [](Net& net, std::string name, nb::dict cfg_dict, std::size_t cap) {
+    /// TRACES: VR-011, VR-002 | DP-001 | PR-002, PR-004
+    ///
+    /// One call builds the whole downstream chain, in the one order that works:
+    ///
+    ///     matcher (fits the calibration)
+    ///       -> registry (needs a discounter built from it)
+    ///         -> tracker (needs both)
+    ///           -> frame_annotation
+    ///             -> result_sink (needs the registry's claims)
+    ///
+    /// This replaces add_face_tracker / add_identity_matcher / add_frame_annotation.
+    /// They were separate because the network is assembled node by node from
+    /// Python -- and that is exactly how the seam broke: the tracker's dependency
+    /// on a calibration that only exists once the matcher is built cannot be
+    /// expressed as three independent factories, so the tracker factory kept
+    /// constructing FaceTrackerFunc{cfg} against a signature that no longer
+    /// existed. A binding that cannot represent the order will eventually be
+    /// called in the wrong one.
+    ///
+    /// DP-001 -- "modes are front-ends and must not fork pipeline logic" -- is
+    /// the requirement this serves. The replay harness is a front-end. Its job is
+    /// to supply frames and read the result, not to re-derive presence.
+    m.def("add_pipeline", [](Net& net, std::string gallery_path, nb::dict cfg_dict,
+                             std::size_t cap, std::string embedder_model,
+                             std::string embedder_sha256) {
         Config cfg = config_from_dict(cfg_dict);
-        auto node = std::make_shared<kpn::ObjectVariantNodeWrapper<
+        cfg.gallery_path = gallery_path;  // so a refreshed calibration persists back
-            FaceTrackerFunc, SaeVariant, kpn::in<"embedded">, kpn::out<"tracked">>>(cap, cfg);
-        net.add(std::move(name), std::move(node));
-    }, "net"_a, "name"_a, "config"_a, "capacity"_a = 16);
 
-    /// TRACES: GR-004 | SR-001
-    // embedder_model / embedder_sha256 identify whatever produced the embeddings
-    // that will be fed in. In a replay those come from the dump's own stamp (see
-    // scripts/optimizer/SCHEMA.md), because there is no live embedder in the
-    // network — the dump *is* the embedder as far as this gallery is concerned.
-    // Passing neither leaves the binding unverifiable, which warns loudly and is
-    // fatal under SAE_REQUIRE_GALLERY_STAMP.
-    m.def("add_identity_matcher", [](Net& net, std::string name, std::string gallery_path,
-                                     nb::dict cfg_dict, std::size_t cap,
-                                     std::string embedder_model,
-                                     std::string embedder_sha256) {
-        Config cfg = config_from_dict(cfg_dict);
-        cfg.gallery_path = gallery_path;  // needed to persist refreshed calibration back
         // Cache loaded galleries by path so a threshold sweep (many networks, same
-        // gallery) pays the ~24s JSON parse only once. The matcher holds a const
+        // gallery) pays the parse once. The matcher holds a const ref; the cache
-        // ref; the cache keeps the gallery alive for the process lifetime.
+        // keeps the gallery alive for the process lifetime.
         static std::map<std::string, std::shared_ptr<ActorGallery>> cache;
         auto it = cache.find(gallery_path);
         if (it == cache.end())
             it = cache.emplace(gallery_path,
                      std::make_shared<ActorGallery>(load_gallery(gallery_path))).first;
 
-        // Checked on every construction, not only on the cache miss: the same
+        /// TRACES: GR-004 | SR-001
-        // process may replay several dumps against one cached gallery.
+        // embedder_model / embedder_sha256 identify whatever produced the
+        // embeddings that will be fed in. In a replay those come from the dump's
+        // own stamp: there is no live embedder here, so the dump *is* the
+        // embedder as far as this gallery is concerned. Checked on every
+        // construction, not only on a cache miss -- one process may replay
+        // several dumps against one cached gallery.
         EmbedderStamp feeding;
         feeding.model_name   = std::move(embedder_model);
         feeding.model_sha256 = std::move(embedder_sha256);
@@ -263,31 +391,101 @@ NB_MODULE(sae_kpn, m) {
                                    : feeding.model_name,
                                cfg.require_gallery_stamp);
 
-        auto node = std::make_shared<kpn::ObjectVariantNodeWrapper<
+        // 1. Matcher first: its constructor fits (or loads) the calibration.
-            IdentityMatcherFunc, SaeVariant, kpn::in<"tracked">, kpn::out<"matched">>>(
+        auto matcher = std::make_shared<MatcherWrap>(cap, *it->second, cfg);
-            cap, *it->second, cfg);
+
-        net.add(std::move(name), std::move(node));
+        // 2. The calibration every other stage must decide in (AR-024).
-    }, "net"_a, "name"_a, "gallery"_a, "config"_a, "capacity"_a = 16,
+        auto same_person = same_person_probability(matcher->functor().calibration());
+
+        // 3. Registry + discounter, from Config (AR-025).
+        TrackRegistry::Config reg_cfg;
+        reg_cfg.track_extinction_sec = cfg.track_extinction_sec;
+        reg_cfg.ownership_logodds    = cfg.ownership_logodds;
+        EvidenceDiscounter::Config disc_cfg;
+        disc_cfg.max_views   = cfg.evidence_max_views;
+        disc_cfg.admit_below = cfg.evidence_admit_below;
+        disc_cfg.rho_max     = cfg.evidence_rho_max;
+        auto registry = std::make_shared<TrackRegistry>(
+            reg_cfg, EvidenceDiscounter(same_person, disc_cfg));
+        matcher->functor().set_registry(registry);
+
+        // 4. Tracker, which needs both.
+        auto tracker = std::make_shared<TrackerWrap>(cap, cfg, registry, same_person);
+
+        // 5. Projection, stateless.
+        auto annot = std::make_shared<AnnotWrap>(cap);
+
+        // 6. The real sink. `done` outlives the network via the session below;
+        //    ResultSinkFunc holds it by reference, as it does in main.cpp.
+        auto session = std::make_shared<ReplaySession>();
+        session->cfg      = cfg;          // the sink holds this by reference
+        session->registry = registry;
+        auto sink = std::make_shared<SinkWrap>(cap, session->cfg, session->done);
+
+        /// TRACES: AR-012, AR-016 | IR-003 | SR-002
+        // The claim path, identical to main.cpp's. Without the flush hook every
+        // track still live at EOF is silently dropped -- which in a replay is
+        // most of the closing scene, and reads as a recognition miss rather than
+        // as a missing wire.
+        ResultSinkFunc& sink_fn = sink->functor();
+        registry->on_track_dead([&sink_fn](const DeadTrack& d) { sink_fn.add_claim(d); });
+        sink_fn.set_pre_write_hook([registry](double last_ts) { registry->flush(last_ts); });
+
+        net.add("tracker", tracker);
+        net.add("matcher", matcher);
+        net.add("annotation", annot);
+        net.add("sink", sink);
+
+        // Keyed by network so release_pipeline can free it. Not a leak-by-design:
+        // a sweep builds one network per replay, and the sink accumulates every
+        // annotation, so holding these forever would grow with films x configs.
+        sessions()[&net] = session;
+    }, "net"_a, "gallery"_a, "config"_a, "capacity"_a = 16,
        "embedder_model"_a = "", "embedder_sha256"_a = "");
 
-    /// TRACES: AR-012, AR-013 | SR-002
+    /// Drop the session for a network. Idempotent. Call after net.stop(); not
-    // Was add_scene_tracker, backed by the extinction-timer state machine. The
+    /// calling it holds one registry and one sink's accumulated frames per
-    // node is gone (see frame_annotation_node.hpp) and so is the timer; this
+    /// replay, which a long sweep will notice.
-    // projects a matched frame into the same SceneAnnotation the Python sink
+    m.def("release_pipeline", [](Net& net) { sessions().erase(&net); }, "net"_a);
-    // already reads, so the seam's output type is unchanged. It takes no config
+
-    // because it has no state to configure -- which is the point.
+    /// TRACES: VR-011 | AR-025 | PR-002
-    m.def("add_frame_annotation", [](Net& net, std::string name, std::size_t cap) {
+    /// The registry's own count of how often it was wrong, exposed so a replay
-        auto node = std::make_shared<kpn::ObjectVariantNodeWrapper<
+    /// can fail on it instead of returning a plausible-looking empty answer.
-            FrameAnnotationFunc, SaeVariant, kpn::in<"matched">, kpn::out<"annotation">>>(cap);
+    ///
-        net.add(std::move(name), std::move(node));
+    /// `dropped_votes` is the one that matters here and it earned its keep
-    }, "net"_a, "name"_a, "capacity"_a = 16);
+    /// immediately. A vote lands on a track the registry has already reaped when
+    /// the matcher lags the tracker by more than track_extinction_sec of film.
+    /// In scene_analyze that cannot happen -- channels are 16-64 deep, so
+    /// backpressure pins the two nodes within a few frames of each other. This
+    /// harness sized every channel to the whole film to avoid a PyNode overflow
+    /// drop, which removed the backpressure entirely: the tracker ran the film
+    /// to the end while the matcher was still in its first minute, every vote
+    /// arrived after its track was gone, no track was ever owned, and the run
+    /// produced zero presence windows while cheerfully reporting 1647 frames
+    /// with an identified face.
+    m.def("pipeline_diagnostics", [](Net& net) {
+        nb::dict d;
+        auto it = sessions().find(&net);
+        if (it == sessions().end() || !it->second->registry) return d;
+        const auto& r = *it->second->registry;
+        d["dropped_votes"]   = r.dropped_votes();
+        d["belief_swaps"]    = r.belief_swaps();
+        d["actor_conflicts"] = r.actor_conflicts();
+        d["live_tracks"]     = static_cast<int>(r.live());
+        return d;
+    }, "net"_a);
+
+    /// True once the sink has written its output. The sink flushes on the EOF
+    /// annotation, so a caller that reads the file before this is racing it.
+    m.def("pipeline_done", [](Net& net) {
+        auto it = sessions().find(&net);
+        return it != sessions().end()
+            && it->second->done.load(std::memory_order_acquire);
+    }, "net"_a);
 
     // ── Runtime setters (persistent-pipeline reuse across a threshold sweep) ─────
     // Build the network once, then change thresholds between replays — no rebuild,
     // no teardown (which is where the ROCm deadlock lives), no gallery reload.
-    using MatcherWrap = kpn::ObjectVariantNodeWrapper<
-        IdentityMatcherFunc, SaeVariant, kpn::in<"tracked">, kpn::out<"matched">>;
-
     m.def("set_prob_threshold", [](Net& net, std::string name, float t) {
         auto* w = dynamic_cast<MatcherWrap*>(net.node_ptr(name));
         if (!w) throw std::runtime_error("set_prob_threshold: '" + name + "' is not an identity_matcher");

src/main.cpp

@@ -71,6 +71,7 @@
 #include "nodes/face_tracker_node.hpp"
 #include "nodes/identity_matcher_node.hpp"
 #include "nodes/frame_annotation_node.hpp"
+#include "inference/scene_detector.hpp"   // ISceneDetector::kWindow, for the join-depth derivation
 #include "nodes/scene_detector_node.hpp"
 #include "scene_boundaries.hpp"
 #include "nodes/scene_boundary_annotator_node.hpp"
@@ -84,8 +85,10 @@
 
 #include <opencv2/core/utility.hpp>   // cv::setNumThreads (SAE_CV_THREADS)
 
+#include <algorithm>
 #include <atomic>
 #include <chrono>
+#include <cmath>
 #include <csignal>
 #include <cstdlib>
 #include <cstring>
@@ -101,12 +104,66 @@
 // ── CLI parsing ───────────────────────────────────────────────────────────────
 
 /// TRACES: AR-010, AR-004 | SR-002
-/// How deeply the sampled branch is buffered behind the dense one. TransNetV2
+/// Depth of the dense branch's own input queue. Part of how far behind the
-/// needs kWindow (100) dense frames before it can score any of them, so the face
+/// fanout head TransNetV2 can be, and therefore an input to the join depth.
-/// branch must lag by at least that much or it asks about frames nobody has
+static constexpr std::size_t kSceneInputDepth = 128;
-/// looked at yet. Backpressure turns depth into lag: the fanout blocks on the
+
-/// slower branch rather than dropping, so the detector simply runs ahead.
+/// TRACES: AR-010, AR-004 | SR-002
-static constexpr std::size_t kSceneJoinDepth = 256;
+/// How far the sampled branch must trail the dense one, in seconds of film.
+///
+/// TransNetV2 needs kWindow (100) dense frames before it can score any of
+/// them, and its input queue can hold kSceneInputDepth more, so in the worst
+/// case it has scored only up to (kSceneInputDepth + kWindow) frames behind
+/// whatever the fanout has just delivered. The face branch must be at least
+/// that far behind, or `scene_annotate` asks about frames nobody has looked at
+/// yet. Backpressure turns depth into lag: the fanout blocks on the slower
+/// branch rather than dropping, so the dense branch simply runs ahead.
+///
+/// Divided by a *lower bound* on native frame rate, because a slower source
+/// makes the same frame count span more film — 24 fps is the floor for the
+/// material this runs on, so it is the conservative choice.
+static constexpr double kMinNativeFps    = 24.0;
+static constexpr double kSceneJoinLagSec =
+    (kSceneInputDepth + ISceneDetector::kWindow) / kMinNativeFps;   // ~9.5 s
+
+/// Margin over that minimum, for jitter in TransNetV2's inference time.
+static constexpr double kSceneJoinSafety = 2.0;
+
+/// TRACES: AR-004 | SR-002
+/// Slots the sampled branch needs to hold `kSceneJoinLagSec` of film.
+///
+/// This used to be a constant 256, which is the whole bug: the requirement is a
+/// span of *film*, and the slots needed to hold it depend on `sample_fps`.
+/// Pinned at 256 it was ~256 s of lag at 1 fps — 27x what the join needs — and
+/// nothing recomputed it if `sample_fps` changed, so the one number the join's
+/// correctness rests on drifted silently with an unrelated knob.
+///
+/// It is also the largest single memory item in the pipeline. Every message
+/// embeds `Frame source`, so a slot on this branch holds a full decoded image:
+/// 256 of them is ~1.5 GB at 1080p, against ~110 MB for the derived depth at
+/// 1 fps. See AR-004 — capacity is counted in items, and only the byte figure
+/// (now correct, see types.hpp) shows what a slot really costs.
+static std::size_t scene_join_depth(float sample_fps) {
+    const double slots = kSceneJoinSafety * kSceneJoinLagSec * sample_fps;
+    // Floor of 16: below that the queue stops absorbing ordinary jitter and
+    // starts throttling the fanout, which would slow the dense branch it
+    // exists to let run ahead.
+    return std::max<std::size_t>(16, static_cast<std::size_t>(std::ceil(slots)));
+}
+
+/// TRACES: AR-004 | SR-002
+/// The decimator's input, on the *full-rate* stream.
+///
+/// This was also kSceneJoinDepth, which put a 256-slot buffer of full-rate
+/// frames in front of the decimator — and at 1 fps against 24 fps native, 23 of
+/// every 24 of those frames exist only to be discarded a moment later. Holding
+/// ~1.5 GB of decoded images for frames the very next node throws away is the
+/// worst available use of the memory budget.
+///
+/// A filter is a pass-through, not a reservoir: the lag belongs *after*
+/// decimation, where a slot buys `1/sample_fps` seconds of film instead of
+/// `1/native_fps`. Sized only to keep the decimator fed.
+static constexpr std::size_t kDecimatorInputDepth = 16;
 
 /// Set when the scene branch is built, so shutdown can report whether the join
 /// actually worked.
@@ -151,6 +208,8 @@ static Config parse_args(int argc, char** argv) {
         else if (arg("--start"))           cfg.start_sec        = std::stod(next());
         else if (arg("--end"))             cfg.end_sec          = std::stod(next());
         else if (arg("--cut-threshold"))   cfg.cut_threshold    = std::stof(next());
+        else if (arg("--presence-mode"))   { std::string m = next(); cfg.presence_mode = (m == "flood") ? PresenceMode::flood : PresenceMode::track_extent; }
+        else if (arg("--scene-xgb-model")) cfg.scene_xgb_model  = next();
         else if (arg("--scene-detect"))         cfg.scene_detect     = true;
         else if (arg("--scene-detector"))       cfg.scene_model      = next();
         else if (arg("--scene-detector-engine")) cfg.scene_engine    = next();
@@ -459,6 +518,45 @@ int main(int argc, char** argv) {
             std::cerr << "\n";
         }
 
+        /// TRACES: AR-025, AR-012 | SR-002
+        // How often the registry was asked about a track it had already reaped.
+        //
+        // A vote is dropped when the matcher lags the tracker by more than
+        // track_extinction_sec of FILM time. The two are adjacent nodes with a
+        // 16-deep channel between them, and the matcher is much the slower of
+        // the pair (a GEMM over the whole gallery against a Hungarian solve over
+        // a handful of boxes), so that channel runs full and the lag is close to
+        // its depth. In frames:
+        //
+        //     lag_sec  ~=  channel_depth / sample_fps
+        //
+        // At the default sample_fps of 1.0 that is ~16 s against a 5 s window,
+        // so votes CAN be dropped here, and each one is identity evidence that
+        // never reached the track it belonged to -- presence under-reported, in
+        // a way that reads as a recognition miss.
+        //
+        // Reported rather than fatal, deliberately, and the distinction from the
+        // dropped-frame case below is real: a dropped frame means the output
+        // describes footage nobody analysed, which is always wrong. A dropped
+        // vote means one observation of a track went missing, which degrades a
+        // claim without falsifying it. There is also no measurement yet of how
+        // often it happens on real content -- so this prints the number that
+        // would justify a harder line rather than presuming it. See VR-017.
+        if (registry) {
+            const int dv = registry->dropped_votes();
+            if (dv > 0) {
+                std::cerr << "[registry] WARNING: " << dv << " identity vote(s) "
+                             "arrived for already-reaped tracks. The matcher is "
+                             "lagging the tracker by more than track_extinction_sec ("
+                          << cfg.track_extinction_sec << "s) of film; presence is "
+                             "under-reported. Raise --track-extinction or reduce the "
+                             "face_tracker/identity_matcher channel depth.\n";
+            }
+            std::cerr << "[registry] belief_swaps=" << registry->belief_swaps()
+                      << "  actor_conflicts=" << registry->actor_conflicts()
+                      << "  dropped_votes=" << dv << "\n";
+        }
+
         bool dropped = false;
         {
             std::lock_guard<std::mutex> lk(event_mtx);
@@ -529,7 +627,7 @@ int main(int argc, char** argv) {
         auto boundaries = std::make_shared<SceneBoundaries>();
         scene_fn.set_boundaries(boundaries);
         kpn::ObjectNode<SceneDetectorFunc, kpn::in<"dense">, kpn::out<>, "scene_detector", 0>
-            scene_node(scene_fn, 128);
+            scene_node(scene_fn, kSceneInputDepth);
 
         // Decimator: keep frames on the sample_fps cadence, drop the rest.
         // eof always passes so downstream shuts down cleanly. Stateful — one
@@ -544,7 +642,7 @@ int main(int argc, char** argv) {
                     return true;
                 }
                 return false;
-            }, kSceneJoinDepth);
+            }, kDecimatorInputDepth);
 
         /// TRACES: AR-010 | SR-002
         // Stamp is_scene_boundary from the detector's published verdict. tol is
@@ -559,7 +657,7 @@ int main(int argc, char** argv) {
         // otherwise look exactly like "no boundary here".
         SceneBoundaryAnnotatorFunc annotate_fn{boundaries, 0.5 / cfg.sample_fps};
         kpn::ObjectNode<SceneBoundaryAnnotatorFunc, kpn::in<"frame">, kpn::out<"frame">,
-                        "scene_annotate", 0> annotate(annotate_fn, kSceneJoinDepth);
+                        "scene_annotate", 0> annotate(annotate_fn, scene_join_depth(cfg.sample_fps));
 
         // Reported at shutdown: without this the join is unverifiable, and an
         // annotator that never fired looks identical to footage with no

src/nodes/camera_position_change_detector_node.hpp

@@ -33,9 +33,29 @@ struct CameraPositionChangeDetectorFunc {
 
     explicit CameraPositionChangeDetectorFunc(const Config& cfg)
         : cut_threshold_(cfg.cut_threshold)
+        , want_rgb_hist_(!cfg.scene_xgb_model.empty())
     {
         std::cerr << "[camera_position_change_detector] cut_threshold="
-                  << cut_threshold_ << "\n";
+                  << cut_threshold_
+                  << (want_rgb_hist_ ? "  (+rgb_hist for scene detector)" : "")
+                  << "\n";
+    }
+
+    // 32-bin-per-channel normalised RGB histogram (96 floats), the exact layout
+    // the XGBoost scene detector was trained on (see embedding_dump_node). Only
+    // computed when a scene model is configured, so it costs nothing otherwise.
+    static std::vector<float> rgb_histogram(const cv::Mat& img) {
+        constexpr int kBins = 32;
+        std::vector<float> out(kBins * 3, 0.f);
+        if (img.empty() || img.channels() != 3) return out;
+        float range[] = {0.f, 256.f}; const float* ranges = range; int bins = kBins;
+        for (int c = 0; c < 3; ++c) {              // OpenCV BGR → store B,G,R blocks
+            cv::Mat h;
+            cv::calcHist(&img, 1, &c, cv::Mat(), h, 1, &bins, &ranges);
+            cv::normalize(h, h, 1.0, 0.0, cv::NORM_L1);
+            for (int b = 0; b < kBins; ++b) out[c*kBins + b] = h.at<float>(b);
+        }
+        return out;
     }
 
     Frame operator()(Frame f) {
@@ -64,11 +84,13 @@ struct CameraPositionChangeDetectorFunc {
         prev_hist_       = hist;
         prev_hist_valid_ = true;
 
+        if (want_rgb_hist_) f.rgb_hist = rgb_histogram(f.image);
         return f;
     }
 
 private:
     float   cut_threshold_;
+    bool    want_rgb_hist_{false};
     cv::Mat prev_hist_;
     bool    prev_hist_valid_{false};
 };

src/nodes/embedding_dump_node.hpp

@@ -5,6 +5,7 @@
 #include "gallery/embedder_stamp.hpp"
 
 #include <H5Cpp.h>
+#include <opencv2/imgproc.hpp>   // cv::calcHist for the per-frame RGB histogram
 
 #include <atomic>
 #include <cstdint>
@@ -167,6 +168,12 @@ struct EmbeddingDumpFunc {
         fidx_.push_back(ef.source.frame_idx);
         is_cut_.push_back(ef.source.is_cut ? 1 : 0);
         is_bnd_.push_back(ef.source.is_scene_boundary ? 1 : 0);
+        // Per-frame normalised RGB histogram (kHistBins per channel), for offline
+        // training of a learned scene-boundary detector against X-Ray scene
+        // boundaries — the grayscale-correlation cut detector is blind on
+        // low-contrast grades (Scarface: 1 cut in 10k frames). Cheap and the frame
+        // is already decoded here; empty frame → zeros.
+        append_rgb_hist(ef.source.image);
         face_off_.push_back(static_cast<int64_t>(conf_.size()));
         face_cnt_.push_back(n);
 
@@ -287,6 +294,11 @@ private:
         write_vec(frames, "is_scene_boundary", is_bnd_, H5::PredType::NATIVE_UINT8);
         write_vec(frames, "face_offset",   face_off_, H5::PredType::NATIVE_INT64);
         write_vec(frames, "face_count",    face_cnt_, H5::PredType::NATIVE_INT32);
+        // Per-frame normalised RGB histogram, kHistBins per channel laid out
+        // [R(kHistBins) G(kHistBins) B(kHistBins)] per row. Feeds the learned
+        // scene-boundary detector (see scripts/scene_detector/).
+        write_vec(frames, "rgb_hist", rgb_hist_, H5::PredType::NATIVE_FLOAT,
+                  kHistBins * 3);
 
         H5::Group faces = file.createGroup("faces");
         write_vec(faces, "embedding",  emb_,  H5::PredType::NATIVE_FLOAT, kEmbedDim);
@@ -301,6 +313,29 @@ private:
                   << conf_.size() << " faces → " << path_ << "\n";
     }
 
+    // Per-channel bin count for the RGB histogram. 32 → a 96-float row per frame,
+    // ~40 KB per 10k-frame film: negligible next to the embeddings.
+    static constexpr int kHistBins = 32;
+
+    // Append the frame's normalised per-channel RGB histogram (R,G,B blocks). An
+    // empty frame (EOF sentinels never reach here) yields a zero row so the array
+    // stays parallel to ts_.
+    void append_rgb_hist(const cv::Mat& img) {
+        const size_t base = rgb_hist_.size();
+        rgb_hist_.resize(base + kHistBins * 3, 0.f);
+        if (img.empty() || img.channels() != 3) return;
+        float range[] = {0.f, 256.f};
+        const float* ranges[] = {range};
+        int bins = kHistBins;
+        for (int c = 0; c < 3; ++c) {                 // OpenCV is BGR; store as B,G,R blocks
+            cv::Mat h;
+            cv::calcHist(&img, 1, &c, cv::Mat(), h, 1, &bins, ranges);
+            cv::normalize(h, h, 1.0, 0.0, cv::NORM_L1);
+            for (int b = 0; b < kHistBins; ++b)
+                rgb_hist_[base + c * kHistBins + b] = h.at<float>(b);
+        }
+    }
+
     std::string   path_, movie_;
     EmbedderStamp stamp_;
     DumpProvenance prov_;
@@ -315,4 +350,5 @@ private:
     std::vector<int32_t> face_cnt_;
     std::vector<float>   emb_, bbox_, lmk_, conf_;
     std::vector<float>   sharp_, resid_;   // AR-028 quality vector, parallel to conf_
+    std::vector<float>   rgb_hist_;        // kHistBins*3 per frame, parallel to ts_
 };

src/nodes/frame_annotation_node.hpp

@@ -38,6 +38,12 @@ struct FrameAnnotationFunc {
 
     SceneAnnotation operator()(MatchedSceneFrame mf) {
         if (mf.source.eof) return {0.0, {}, /*eof=*/true};
-        return {mf.source.timestamp_sec, std::move(mf.actors)};
+        SceneAnnotation sa;
+        sa.timestamp_sec     = mf.source.timestamp_sec;
+        sa.visible_actors    = std::move(mf.actors);
+        sa.is_cut            = mf.source.is_cut;
+        sa.is_scene_boundary = mf.source.is_scene_boundary;
+        sa.rgb_hist          = std::move(mf.source.rgb_hist);
+        return sa;
     }
 };

src/nodes/identity_matcher_node.hpp

@@ -152,7 +152,12 @@ struct IdentityMatcherFunc {
     /// Where per-frame identity evidence reaches the registry. Optional: with no
     /// registry attached the matcher behaves exactly as before, which keeps the
     /// replay harness and the unit tests working unchanged.
-    void set_registry(std::shared_ptr<TrackRegistry> r) { registry_ = std::move(r); }
+    void set_registry(std::shared_ptr<TrackRegistry> r) {
+        registry_ = std::move(r);
+        // This node is the evidence source, so the registry must not close a
+        // track until this node's watermark has passed it (AR-013).
+        if (registry_) registry_->expect_evidence();
+    }
 
     // Runtime setter — lets a persistent pipeline be reused across a threshold sweep
     // without rebuilding the (expensive, gallery-resident) matcher. The gallery,
@@ -165,6 +170,22 @@ struct IdentityMatcherFunc {
             return {std::move(tf.source), {}};
         }
 
+        /// TRACES: AR-012, AR-013 | SR-002
+        // Publish the evidence watermark BEFORE voting on this frame: every
+        // observation strictly before it has now been folded in, so the registry
+        // may reap against it. Unconditional -- a frame with no faces still
+        // advances the watermark, or a long faceless stretch would stall reaping
+        // and hold every dormant track open to the end of the film.
+        //
+        // This is what makes presence independent of node speed. The registry
+        // used to reap on the TRACKER's clock, and backpressure (working as
+        // AR-004 intends) means the tracker can be a whole channel's depth ahead
+        // of this node -- so tracks were closed before their votes arrived, the
+        // votes were dropped, and the run silently under-reported. Measured on
+        // the SuperHero fixture before this change: channel depth 32 gave 5
+        // actors, depth 10322 gave 0, from identical input.
+        if (registry_) registry_->advance_evidence(tf.source.timestamp_sec);
+
         // A hard cut changes the camera viewpoint. The face_tracker may revive a
         // track_id across the cut (identity continuity), but promotion must never
         // mix embeddings from two viewpoints under one buffer, so we still drop

src/nodes/result_sink_node.hpp

@@ -3,6 +3,10 @@
 #include "types.hpp"
 #include "config.hpp"
 #include "track_registry.hpp"
+#ifdef SAE_SCENE_XGB
+#include "inference/xgb_scene_boundary.hpp"
+#include "inference/audio_logpsd.hpp"
+#endif
 
 #include <nlohmann/json.hpp>
 #include <algorithm>
@@ -184,6 +188,20 @@ private:
             aw.scenes.push_back({c.first_seen, c.last_seen, c.belief, c.route});
         }
 
+        // Flood-fill: snap each claim to the shot it sits in, so an actor seen
+        // once in a scene is reported across the whole scene. Bounded by real
+        // TransNetV2 boundaries — a window never crosses one — and a no-op when
+        // scene detection found no boundaries (nothing to snap to).
+        if (cfg_.presence_mode == PresenceMode::flood) {
+            const std::vector<double> bounds = scene_boundaries();
+            if (!bounds.empty())
+                for (auto& [idx, aw] : by_actor)
+                    for (auto& w : aw.scenes) {
+                        w.start = boundary_at_or_before(bounds, w.start);
+                        w.end   = boundary_after(bounds, w.end);
+                    }
+        }
+
         std::vector<ActorWindow> result;
         for (auto& [idx, aw] : by_actor) {
             std::sort(aw.scenes.begin(), aw.scenes.end(),
@@ -193,6 +211,90 @@ private:
         return result;
     }
 
+    // Sorted, de-duplicated boundary timestamps seen this run, framed by the
+    // film's own extent so the first and last shots are closed intervals. Derived
+    // from frames_ rather than a separate accumulator: the frames are already
+    // retained and this runs once.
+    //
+    // Prefers TransNetV2 shot boundaries (is_scene_boundary) when a scene detector
+    // populated them; otherwise falls back to the always-on histogram cuts
+    // (is_cut, camera_position_change_detector). On this ROCm box the scene
+    // detector cannot run in-process (see the dumper note), so is_cut is what
+    // flood-fill actually snaps to — coarser than true shot boundaries (cuts also
+    // fire on in-shot angle changes) but present with no extra pass.
+    std::vector<double> scene_boundaries() const {
+        std::vector<double> b;
+        b.push_back(0.0);
+
+        // Preferred: the learned XGBoost scene detector, run once here post-EOF
+        // (the knee threshold needs the whole film, so this is inherently a final
+        // step — like flood-fill itself). Measured best flood boundary source.
+        std::vector<double> learned = xgb_boundaries();
+        if (!learned.empty()) {
+            for (double t : learned) b.push_back(t);
+        } else {
+            // Fallback: TransNetV2 shot boundaries if present, else histogram cuts.
+            bool have_scene = false;
+            for (const auto& sa : frames_)
+                if (sa.is_scene_boundary) { have_scene = true; break; }
+            for (const auto& sa : frames_) {
+                const bool boundary = have_scene ? sa.is_scene_boundary : sa.is_cut;
+                if (boundary) b.push_back(sa.timestamp_sec);
+            }
+        }
+        b.push_back(last_ts_ + 1.0);   // a right edge past the final sample
+        std::sort(b.begin(), b.end());
+        b.erase(std::unique(b.begin(), b.end()), b.end());
+        return b;
+    }
+
+    // Run the learned scene-boundary detector over the collected per-frame RGB
+    // histograms + per-second audio log-PSD (decoded once from the movie). Returns
+    // {} when no model is configured, the build lacks XGBoost, or no rgb_hist was
+    // stamped (camera-position node only does so when a model is set).
+    std::vector<double> xgb_boundaries() const {
+#ifdef SAE_SCENE_XGB
+        if (cfg_.scene_xgb_model.empty()) return {};
+        std::vector<std::vector<float>> hist;
+        std::vector<double> ts;
+        hist.reserve(frames_.size()); ts.reserve(frames_.size());
+        for (const auto& sa : frames_) {
+            if (sa.rgb_hist.empty()) return {};   // hist not stamped → bail to fallback
+            hist.push_back(sa.rgb_hist);
+            ts.push_back(sa.timestamp_sec);
+        }
+        if (hist.size() < 16) return {};
+        try {
+            auto audio = AudioLogPSD::extract(cfg_.movie_path);   // [T'][B], aligned per second
+            if ((int)audio.size() != (int)hist.size())
+                audio.resize(hist.size(),
+                             std::vector<float>(audio.empty() ? 57 : audio[0].size(), 0.f));
+            XGBSceneBoundary det(cfg_.scene_xgb_model);
+            auto b = det.boundaries(hist, ts, audio);
+            std::cerr << "[result_sink] XGBoost scene detector: " << b.size()
+                      << " boundaries\n";
+            return b;
+        } catch (const std::exception& e) {
+            std::cerr << "[result_sink] scene detector failed (" << e.what()
+                      << "), falling back to histogram cuts\n";
+            return {};
+        }
+#else
+        return {};
+#endif
+    }
+
+    // The boundary opening the shot that contains t (largest boundary ≤ t).
+    static double boundary_at_or_before(const std::vector<double>& b, double t) {
+        auto it = std::upper_bound(b.begin(), b.end(), t);
+        return (it == b.begin()) ? b.front() : *(it - 1);
+    }
+    // The boundary closing the shot that contains t (smallest boundary > t).
+    static double boundary_after(const std::vector<double>& b, double t) {
+        auto it = std::upper_bound(b.begin(), b.end(), t);
+        return (it == b.end()) ? b.back() : *it;
+    }
+
     json build_epochs() {
         json actors = json::array();
         for (const auto& aw : build_actor_windows()) {

src/nodes/scene_detector_node.hpp

@@ -6,6 +6,8 @@
 #include <memory>
 #include "inference/scene_detector.hpp"
 
+#include <opencv2/imgproc.hpp>   // cv::resize, for to_model_input
+
 #include <nlohmann/json.hpp>
 #include <algorithm>
 #include <atomic>
@@ -76,7 +78,7 @@ struct SceneDetectorFunc {
         }
         prev_ts_ = f.timestamp_sec;
 
-        images_.push_back(f.image);
+        images_.push_back(to_model_input(f.image));
         times_.push_back(f.timestamp_sec);
 
         // Once we have a full window, score it and slide forward by `stride`.
@@ -90,6 +92,48 @@ struct SceneDetectorFunc {
         }
     }
 
+    /// TRACES: AR-004, AR-010 | SR-002
+    /// Reduce a decoded frame to exactly what TransNetV2 consumes, once.
+    ///
+    /// The window used to hold the frames as decoded — full resolution — and
+    /// leave the downscale to the backend. But the model's input is 48x27
+    /// (`ISceneDetector::kFrameW/H`; the config note for `dense_scale` says so
+    /// outright: "TransNetV2 downsamples to 48x27 regardless"), so the buffer
+    /// held ~590 MB at 1080p to feed something that needs ~380 KB. That is not
+    /// a channel capacity, so no amount of tuning channel depths would ever
+    /// have found it.
+    ///
+    /// It is also redundant work. Windows overlap by `kWindow - stride`, so a
+    /// frame appears in several of them and was re-downscaled once per window;
+    /// now it is downscaled once, when it arrives.
+    ///
+    /// **This must reproduce the backends' preprocessing exactly**, because the
+    /// project invariant is that every model gets the input it was trained for
+    /// — a model run off-distribution returns confident, plausible, wrong
+    /// output, and here that means fabricated shot boundaries. Both
+    /// ort_backend.cpp and trt_backend.cpp guard mis-sized input with, in this
+    /// order, `convertTo(CV_8UC3)` then
+    /// `cv::resize(..., {kFrameW, kFrameH}, 0, 0, cv::INTER_AREA)`. The same
+    /// two operations are done here, so the tensor the model receives is
+    /// unchanged; the backend guard then sees a correctly-sized frame and does
+    /// nothing. The interface has always specified this shape as the caller's
+    /// job ("Each frame must already be kFrameW x kFrameH, BGR, CV_8UC3"), so
+    /// this makes the node meet a contract it was already given.
+    static cv::Mat to_model_input(const cv::Mat& src) {
+        cv::Mat typed;
+        if (src.type() != CV_8UC3) src.convertTo(typed, CV_8UC3);
+        else                       typed = src;
+
+        if (typed.cols == ISceneDetector::kFrameW &&
+            typed.rows == ISceneDetector::kFrameH)
+            return typed;
+
+        cv::Mat small;
+        cv::resize(typed, small, {ISceneDetector::kFrameW, ISceneDetector::kFrameH},
+                   0, 0, cv::INTER_AREA);
+        return small;
+    }
+
     /// TRACES: AR-011 | SR-002
     // How close two boundaries have to be before they are the same boundary,
     // derived from the cadence the detector was actually fed.

src/tools/scene_features_dump.cpp

@@ -0,0 +1,58 @@
+// scene_features_dump — write the C++ scene-boundary feature matrix to HDF5, so
+// the XGBoost model is TRAINED on exactly the features the C++ detector produces
+// at inference (parity by construction — no numpy re-implementation to keep in
+// sync). Reads frames/rgb_hist + frames/timestamp_sec from a dump and, given the
+// movie, the per-second audio log-PSD; writes features [T,206] + timestamps.
+//
+//   scene_features_dump <dump.h5> <movie> <out_features.h5>
+//
+// The py3.12 venv trainer (train_xgb_cpp.py) reads <out_features.h5>, attaches
+// the soft Gaussian boundary target, fits XGBoost, and saves the model that
+// XGBSceneBoundary loads. Same C++ features both sides → exact parity.
+
+#include "inference/xgb_scene_boundary.hpp"
+#include "inference/audio_logpsd.hpp"
+#include <H5Cpp.h>
+#include <iostream>
+#include <vector>
+
+int main(int argc, char** argv) {
+    if (argc < 4) {
+        std::cerr << "usage: scene_features_dump <dump.h5> <movie> <out.h5>\n";
+        return 1;
+    }
+    H5::H5File in(argv[1], H5F_ACC_RDONLY);
+    H5::DataSet hd = in.openDataSet("frames/rgb_hist");
+    hsize_t hdims[2]; hd.getSpace().getSimpleExtentDims(hdims);
+    std::vector<float> flat(hdims[0]*hdims[1]);
+    hd.read(flat.data(), H5::PredType::NATIVE_FLOAT);
+    const int T = hdims[0], C = hdims[1];
+    std::vector<std::vector<float>> hist(T, std::vector<float>(C));
+    for (int t = 0; t < T; ++t)
+        for (int c = 0; c < C; ++c) hist[t][c] = flat[t*C+c];
+
+    H5::DataSet td = in.openDataSet("frames/timestamp_sec");
+    hsize_t tdim[1]; td.getSpace().getSimpleExtentDims(tdim);
+    std::vector<double> ts(tdim[0]);
+    td.read(ts.data(), H5::PredType::NATIVE_DOUBLE);
+
+    auto audio = AudioLogPSD::extract(argv[2]);
+    if ((int)audio.size() != T) {
+        std::cerr << "[features] audio rows " << audio.size() << " != hist rows "
+                  << T << " — aligning (pad/truncate)\n";
+        audio.resize(T, std::vector<float>(audio.empty()?57:audio[0].size(), 0.f));
+    }
+
+    std::vector<float> X = XGBSceneBoundary::feature_matrix(hist, audio);
+    const int F = XGBSceneBoundary::kNFeatures;
+
+    H5::H5File out(argv[3], H5F_ACC_TRUNC);
+    hsize_t xd[2] = {(hsize_t)T, (hsize_t)F};
+    out.createDataSet("features", H5::PredType::NATIVE_FLOAT, H5::DataSpace(2, xd))
+       .write(X.data(), H5::PredType::NATIVE_FLOAT);
+    hsize_t td2[1] = {(hsize_t)T};
+    out.createDataSet("timestamp_sec", H5::PredType::NATIVE_DOUBLE, H5::DataSpace(1, td2))
+       .write(ts.data(), H5::PredType::NATIVE_DOUBLE);
+    std::cerr << "[features] wrote [" << T << "," << F << "] → " << argv[3] << "\n";
+    return 0;
+}

src/tools/xgb_boundary_parity.cpp

@@ -0,0 +1,82 @@
+// Parity harness: run the C++ XGBSceneBoundary on a dump's frames/rgb_hist and
+// print the boundary timestamps, so they can be diffed against the Python
+// knee_boundaries (scripts/scene_detector). Feature parity is the whole risk of
+// the C++ port; this proves it before wiring into the pipeline.
+//
+//   xgb_boundary_parity <dump.h5> <model.json>
+//
+// Prints: "<n> boundaries: t0 t1 t2 ..."
+
+#include "inference/xgb_scene_boundary.hpp"
+#include "inference/audio_logpsd.hpp"
+#include <H5Cpp.h>
+#include <iostream>
+#include <fstream>
+#include <string>
+#include <vector>
+
+int main(int argc, char** argv) {
+    if (argc < 3) { std::cerr << "usage: xgb_boundary_parity <dump.h5> <model.json>\n"; return 1; }
+    H5::H5File f(argv[1], H5F_ACC_RDONLY);
+
+    auto read2d = [&](const char* name, std::vector<std::vector<float>>& out, int cols) {
+        H5::DataSet ds = f.openDataSet(name);
+        H5::DataSpace sp = ds.getSpace();
+        hsize_t dims[2]; sp.getSimpleExtentDims(dims);
+        std::vector<float> flat(dims[0]*dims[1]);
+        ds.read(flat.data(), H5::PredType::NATIVE_FLOAT);
+        out.assign(dims[0], std::vector<float>(cols));
+        for (hsize_t i = 0; i < dims[0]; ++i)
+            for (int j = 0; j < cols; ++j) out[i][j] = flat[i*dims[1]+j];
+    };
+    std::vector<std::vector<float>> hist;
+    read2d("frames/rgb_hist", hist, XGBSceneBoundary::kHistBins*3);
+
+    H5::DataSet tsd = f.openDataSet("frames/timestamp_sec");
+    hsize_t td[1]; tsd.getSpace().getSimpleExtentDims(td);
+    std::vector<double> ts(td[0]);
+    tsd.read(ts.data(), H5::PredType::NATIVE_DOUBLE);
+
+    // parity debug: print video features for row 100 (compare to Python)
+    if (argc > 3 && std::string(argv[3]) == "--row100") {
+        auto base = XGBSceneBoundary::debug_base(hist, {});
+        std::cout << "row100:";
+        for (int j = 0; j < 17; ++j) std::cout << " " << base[100][j];
+        std::cout << "\n";
+        return 0;
+    }
+
+    // --feat <cpp_features.h5>: predict directly on the dumped C++ feature matrix
+    // (same bytes Python reads) — a clean parity check with no live-decode variance.
+    if (argc > 4 && std::string(argv[3]) == "--feat") {
+        H5::H5File ff(argv[4], H5F_ACC_RDONLY);
+        H5::DataSet fd = ff.openDataSet("features");
+        hsize_t fdm[2]; fd.getSpace().getSimpleExtentDims(fdm);
+        std::vector<float> X(fdm[0]*fdm[1]);
+        fd.read(X.data(), H5::PredType::NATIVE_FLOAT);
+        XGBSceneBoundary det(argv[2]);
+        auto pdbg = det.debug_predict(X, int(fdm[0]), int(fdm[1]));
+        auto pk = XGBSceneBoundary::debug_find_peaks(pdbg, 5);
+        std::cerr << "[parity] C++ raw peaks=" << pk.size() << "\n";
+        auto b = det.boundaries_from_features(X, int(fdm[0]), int(fdm[1]), ts);
+        std::cout << b.size() << " boundaries:";
+        for (double t : b) std::cout << " " << int(t);
+        std::cout << "\n";
+        return 0;
+    }
+
+    // Optional movie path (argv[4]): decode audio → per-second log-PSD.
+    std::vector<std::vector<float>> audio;
+    if (argc > 4) {
+        audio = AudioLogPSD::extract(argv[4]);
+        std::cerr << "[parity] audio rows=" << audio.size()
+                  << " (hist rows=" << hist.size() << ")\n";
+    }
+
+    XGBSceneBoundary det(argv[2]);
+    auto b = det.boundaries(hist, ts, audio);
+    std::cout << b.size() << " boundaries:";
+    for (double t : b) std::cout << " " << int(t);
+    std::cout << "\n";
+    return 0;
+}

src/track_registry.hpp

@@ -97,7 +97,12 @@ struct Track {
     Embedding             mean{};                 ///< running directional mean
     std::vector<Embedding> views;                 ///< distinct looks, for AR-025 discounting
     float                 discounted_weight{0.f}; ///< sum of applied weights
-    int                   n_obs{0};
+    int                   n_obs{0};                ///< every scored face on this track
+    /// Observations that were actually evidence, and so spent the correlation
+    /// budget. Indexing the effective-sample correction by this rather than by
+    /// n_obs is what stops non-matches exhausting it — see
+    /// Config::evidence_floor_p.
+    int                   n_evidence{0};
 
     bool on_screen() const { return !last_seen.has_value(); }
 };
@@ -119,6 +124,39 @@ public:
         /// extends a presence claim.
         double track_extinction_sec{5.0};
         float  ownership_logodds{2.0f};   ///< belief needed to own a track (~0.88 posterior)
+
+        /// TRACES: AR-025 | SR-002
+        /// Posterior below which an observation is not evidence *for* an actor,
+        /// and so does not spend that actor's correlation budget.
+        ///
+        /// The budget is an effective-sample correction: with observations
+        /// correlated at rho, the weight of the n-th is
+        /// `n_eff(n+1) - n_eff(n) = 2/((n+1)(n+2))` at rho=0.5, so it decays
+        /// quadratically and the total converges to 1/rho = 2. That is the
+        /// intended behaviour — a long static shot must not out-argue varied
+        /// evidence purely by lasting longer.
+        ///
+        /// What was not intended is *who spends it*. Every scored face was
+        /// folded in, so an observation at p=0.02 — which contributes
+        /// log(0.98) = -0.02 of belief, nothing — consumed the same increment
+        /// as one at p=0.95. On SuperHero-2 track 3 that exhausted the budget
+        /// on the frames that recognised nobody: 103 observations, effective
+        /// weight 2.026, belief 0.455 against a 0.881 threshold, with the 51
+        /// frames that did identify the actor arriving when each was worth
+        /// 0.0002. The identification was lost.
+        ///
+        /// It also made the answer depend on frame rate, which is the defect
+        /// AR-013 already had to fix once: deliver more frames, dilute the
+        /// budget with more non-matches, and a track that was owned stops
+        /// being owned. Measured — the same clip identified the actor before a
+        /// KPN throughput fix and not after, from identical input.
+        ///
+        /// 0.5 is the point where the posterior stops favouring the hypothesis
+        /// at all, not a tuned threshold. Near-misses still count, which is the
+        /// design: an observation at 0.6 is evidence and is folded in. Below
+        /// 0.5 the observation argues *against*, which noisy-OR cannot
+        /// represent, so nothing is lost by declining to spend a budget on it.
+        float  evidence_floor_p{0.5f};
     };
 
     /// The discounter is a constructor argument rather than an option: there is
@@ -138,13 +176,65 @@ public:
         FrameScope(TrackRegistry& reg, double now)
             : reg_(reg), lock_(reg.mu_) { reg_.tick_locked(now); }
 
-        /// All live tracks — **one pool**. `last_seen` tells the caller whether
+        /// All ASSOCIABLE tracks — **one pool**. `last_seen` tells the caller
-        /// IoU is meaningful; a dormant track is matched on embedding alone.
+        /// whether IoU is meaningful; a dormant track is matched on embedding
-        /// There is no separate revival path (AR-008).
+        /// alone. There is no separate revival path (AR-008).
+        ///
+        /// TRACES: AR-008, AR-013 | SR-002
+        /// Association and reaping share ONE clock — the evidence watermark when a
+        /// matcher is attached, the tracker clock otherwise (they coincide when
+        /// there is only one). `candidates()` and `reap_locked()` apply the SAME
+        /// `track_extinction_sec` horizon against that clock, so the offered pool
+        /// and the live pool are the same set:
+        ///
+        ///   offered      ⟺ (clock - last_seen) ≤ track_extinction_sec
+        ///   reaped/erased ⟺ (clock - last_seen) >  track_extinction_sec
+        ///
+        /// This closes two symmetric failures. (1) Offering on the tracker's clock
+        /// (ahead of the watermark) let a face associate onto a track the registry
+        /// had ALREADY reaped on the watermark; the vote then landed on a dead id
+        /// and was dropped (record_vote → dropped_votes_). Rare live (small lag),
+        /// but replay runs the tracker far ahead of the matcher and lost ~0.3% of
+        /// votes. (2) Historically, offering on a LOOSER horizon than the reap left
+        /// retired tracks in the pool while the matcher lagged, so a new face
+        /// re-associated onto a long-dead track and two people merged into one
+        /// window (measured: 5 actors/16 windows at depth 32 vs 3/5 at depth 10322).
+        /// A single clock and a single threshold make both impossible: nothing is
+        /// offered past its reap horizon, nothing is reaped while still offerable.
         std::vector<Track*> candidates() {
             std::vector<Track*> out;
             out.reserve(reg_.tracks_.size());
-            for (auto& [id, t] : reg_.tracks_) out.push_back(&t);
+            // Filter association on the SAME clock reaping uses (the evidence
+            // watermark when a matcher is attached, else the tracker clock). The
+            // two used to differ deliberately — the tracker offered on now_ while
+            // the registry reaped on evidence_through_ — but that let the tracker
+            // associate a face onto a track the registry had already reaped on the
+            // watermark, whose vote then landed on a dead id and was dropped
+            // (record_vote → dropped_votes_). In the live pipeline the lag is tiny
+            // so it rarely bit; in replay the Python source runs the tracker far
+            // ahead of the matcher and ~0.3% of votes were lost. One clock for both
+            // "may this associate?" and "is this reaped?" closes the race: a track
+            // past the horizon is neither offered nor reaped-out-from-under a vote.
+            const double clock =
+                reg_.awaits_evidence_ ? reg_.evidence_through_ : reg_.now_;
+            for (auto& [id, t] : reg_.tracks_) {
+                // On-screen tracks are always candidates (actively tracked this
+                // frame). A dormant (off-screen) track is only worth keeping alive
+                // for re-association if it was actually IDENTIFIED: an unowned
+                // dormant track has no actor to re-attach to, so holding it in the
+                // pool only bloats the matcher's per-frame comparison set (every
+                // candidate is a GEMM row) and invites a new face re-associating
+                // onto an anonymous stub. Gating dormant tracks on t.actor keeps
+                // the pool bounded regardless of how large track_extinction_sec is
+                // — which is what makes a long re-association window affordable.
+                if (t.last_seen) {   // dormant
+                    if (!t.actor.has_value())
+                        continue;    // never identified: not worth re-associating
+                    if ((clock - *t.last_seen) > reg_.cfg_.track_extinction_sec)
+                        continue;    // past the re-association horizon
+                }
+                out.push_back(&t);
+            }
             return out;
         }
 
@@ -164,6 +254,50 @@ public:
     /// happens to appear, and a film ending mid-track never closes.
     void tick(double now) { std::lock_guard g(mu_); tick_locked(now); }
 
+    /// TRACES: AR-012, AR-013, AR-025 | SR-002
+    /// The evidence watermark: every observation up to `t` has been folded in.
+    ///
+    /// Reaping is driven by THIS, not by the tracker's clock, and the difference
+    /// is what stops a correct answer from depending on how fast two nodes run.
+    ///
+    /// The tracker and the matcher are separate KPN nodes with a channel between
+    /// them, and the matcher is much the slower of the pair. Backpressure —
+    /// working exactly as AR-004 intends — turns that channel's depth into lag,
+    /// so the tracker's timestamp can be far ahead of the last frame anybody has
+    /// actually voted on. Reaping on the tracker's clock therefore closed tracks
+    /// before their evidence arrived: the votes landed on ids that no longer
+    /// existed, were counted as dropped, and the track was emitted unowned or
+    /// not at all. Deeper channel, fewer identifications, from identical input.
+    ///
+    /// The fix is not to bound the channel against `track_extinction_sec`. That
+    /// makes an algorithm constant police a throughput knob, and leaves the
+    /// answer a function of scheduling. It is to reap on the watermark, which is
+    /// the same device `SceneBoundaries::scored_through()` uses for the AR-010
+    /// join: a consumer past that point is asking about frames nobody has looked
+    /// at yet, and the honest response is to wait rather than to guess.
+    ///
+    /// Monotonic, and only ever *delays* a reap, so no window can be extended by
+    /// it — AR-013's "a window ends at the last sighting, never after" is a
+    /// property of `emit_locked`, which takes `last_seen` and never `now`.
+    void advance_evidence(double t) {
+        std::lock_guard g(mu_);
+        if (t > evidence_through_) evidence_through_ = t;
+        reap_locked();
+    }
+
+    /// TRACES: AR-013, AR-025 | SR-002
+    /// Declare that some stage will publish an evidence watermark, so reaping
+    /// must wait for it.
+    ///
+    /// Explicit rather than inferred from "has anyone voted yet". Inferring it
+    /// re-opens the bug exactly at startup: before the matcher's first frame no
+    /// vote has been seen, so the registry would fall back to the tracker's
+    /// clock during precisely the window in which the tracker is furthest
+    /// ahead. `IdentityMatcherFunc::set_registry` calls this, so any pipeline
+    /// with a matcher waits, and a test that drives the tracker alone keeps the
+    /// simple behaviour instead of hanging on a watermark nobody will publish.
+    void expect_evidence() { std::lock_guard g(mu_); awaits_evidence_ = true; }
+
     // ── Evidence ─────────────────────────────────────────────────────────────
     /// Fold one observation into a track's belief (AR-025).
     ///
@@ -188,7 +322,13 @@ public:
         if (it == tracks_.end()) { ++dropped_votes_; return; }
 
         Track& t = it->second;
-        const float w = discounter_.weight(t.views, t.n_obs, e);
+        ++t.n_obs;   // every scored face is seen, whether or not it is evidence
+
+        // Not evidence *for* this actor: contributes ~nothing to the belief and
+        // must not spend the correlation budget. See Config::evidence_floor_p.
+        if (posterior < cfg_.evidence_floor_p) return;
+
+        const float w = discounter_.weight(t.views, t.n_evidence, e);
 
         // Weighted lazy-OR: P_new = 1 − (1 − P_old)·(1 − p)^w, which in log
         // space is a plain sum. w is the discounted evidence (AR-025), so a
@@ -197,7 +337,7 @@ public:
         const float p = std::min(1.f - 1e-6f, std::max(1e-6f, posterior));
         t.belief[actor_idx] += w * std::log(1.f - p);
         t.discounted_weight += w;
-        ++t.n_obs;
+        ++t.n_evidence;
 
         const int   best = argmax_belief(t);
         const float best_p = 1.f - std::exp(t.belief[best]);
@@ -259,9 +399,20 @@ public:
 private:
     // ── Locked internals ─────────────────────────────────────────────────────
     void tick_locked(double now) {
+        // The tracker's clock still bounds association (a dormant track is only
+        // a candidate while it is alive), but it no longer decides death.
+        now_ = now;
+        reap_locked();
+    }
+
+    /// Reap against the evidence watermark when a producer of one is attached
+    /// (see expect_evidence); otherwise against the tracker's clock, which is
+    /// the same thing when there is only one clock.
+    void reap_locked() {
+        const double clock = awaits_evidence_ ? evidence_through_ : now_;
         for (auto it = tracks_.begin(); it != tracks_.end(); ) {
             const auto& ls = it->second.last_seen;
-            if (ls && (now - *ls) > cfg_.track_extinction_sec) {
+            if (ls && (clock - *ls) > cfg_.track_extinction_sec) {
                 emit_locked(it->second, *ls);
                 it = tracks_.erase(it);
             } else {
@@ -383,6 +534,9 @@ private:
     std::map<int, int>    owner_index_;   ///< actor_idx → live track_id (AR-015)
     DeadTrackFn           on_dead_;
     int                   next_id_{0};
+    double                now_{0.0};              ///< tracker's clock (association)
+    double                evidence_through_{0.0}; ///< matcher's watermark (reaping)
+    bool                  awaits_evidence_{false};
     int                   dropped_votes_{0};
     int                   belief_swaps_{0};
     int                   actor_conflicts_{0};

src/types.hpp

@@ -31,6 +31,10 @@ struct Frame {
     float    cut_score{0.f};  // histogram cut score = 1 - hist_corr (0=identical, ~1=cut); HUD/debug
     float    bbox_upscale{1.f};  // multiply detector bboxes/landmarks by this to map back to
                                  // original video resolution (>1 when dense_scale downscaled the frame)
+    // Normalised 32-bin-per-channel RGB histogram (96 floats), stamped by the
+    // camera-position node and carried to the sink for the learned scene-boundary
+    // detector (post-EOF, flood-fill boundaries). Empty when scene detection off.
+    std::vector<float> rgb_hist;
 };
 
 // ── CutEvent ──────────────────────────────────────────────────────────────────
@@ -155,6 +159,16 @@ struct SceneAnnotation {
     double                       timestamp_sec{0.0};
     std::vector<IdentifiedActor> visible_actors;
     bool                         eof{false};
+    // Carried through from Frame so the sink can collect boundaries for flood-fill
+    // presence (PresenceMode::flood). is_cut is the always-on histogram cut
+    // (camera_position_change_detector) — the boundary flood-fill uses by default.
+    // is_scene_boundary is the opt-in TransNetV2 shot boundary (0 unless scene
+    // detection ran); kept for a future out-of-process scene detector.
+    bool                         is_cut{false};
+    bool                         is_scene_boundary{false};
+    // Per-frame RGB histogram, carried to the sink for the learned scene-boundary
+    // detector run post-EOF (flood-fill). Empty unless scene detection is enabled.
+    std::vector<float>           rgb_hist;
 };
 
 // ── Actor gallery ─────────────────────────────────────────────────────────────
@@ -184,3 +198,111 @@ struct ActorGallery {
     bool     calib_valid{false};
     uint64_t calib_hash{0};
 };
+
+// ── Channel byte accounting ───────────────────────────────────────────────────
+/// TRACES: AR-004 | SR-002
+///
+/// KPN measures a channel's occupancy in *items* and its bandwidth in bytes,
+/// and gets the byte figure from `kpn::ChannelDataSize<T>`. That primary
+/// template returns `sizeof(T)` — right for a POD, badly wrong for every type
+/// below, each of which is a handful of vectors and a `cv::Mat` header owning
+/// megabytes on the heap.
+///
+/// Unspecialised, the diagnostics reported roughly 200 bytes for a message
+/// carrying a full decoded frame — off by four orders of magnitude at 1080p.
+/// That is not merely a cosmetic stat: it is the one instrument for choosing
+/// channel capacities against a memory ceiling, which is the open half of
+/// AR-004, and it was reading fiction.
+///
+/// **What the number means.** `cv::Mat` is reference-counted, so one decoded
+/// frame referenced from several messages is counted once per reference. The
+/// sum is therefore an upper bound on distinct bytes, and the right bound for
+/// the question being asked: how much would this channel keep alive if nothing
+/// else held it.
+///
+/// Declared against a forward declaration rather than including
+/// `<kpn/channel.hpp>` here, so the message definitions keep no dependency on
+/// the framework that carries them — and so any translation unit that can see
+/// these types also sees their sizes, which is what stops one channel being
+/// instantiated with the default and another with the specialisation.
+
+namespace kpn { template<typename T> struct ChannelDataSize; }
+
+namespace sae::bytes {
+
+inline std::size_t of(const cv::Mat& m) {
+    return m.empty() ? 0u : m.total() * m.elemSize();
+}
+inline std::size_t of(const std::vector<cv::Mat>& v) {
+    std::size_t n = 0;
+    for (const auto& m : v) n += of(m);
+    return n;
+}
+inline std::size_t of(const Frame& f) { return sizeof(Frame) + of(f.image); }
+
+inline std::size_t of(const std::vector<IdentifiedActor>& v) {
+    std::size_t n = v.size() * sizeof(IdentifiedActor);
+    for (const auto& a : v) {
+        n += of(a.crop);
+        // The id strings are short but there is one set per actor per frame,
+        // and a crowd frame carries dozens.
+        n += a.name.capacity() + a.imdb_id.capacity()
+           + a.tmdb_id.capacity() + a.jellyfin_id.capacity();
+    }
+    return n;
+}
+
+}  // namespace sae::bytes
+
+template<> struct kpn::ChannelDataSize<Frame> {
+    static std::size_t bytes(const Frame& f) { return sae::bytes::of(f); }
+};
+
+template<> struct kpn::ChannelDataSize<SceneFrame> {
+    static std::size_t bytes(const SceneFrame& v) {
+        return sizeof(SceneFrame) + sae::bytes::of(v.source)
+             + v.faces.size() * sizeof(DetectedFace);
+    }
+};
+
+template<> struct kpn::ChannelDataSize<AlignedSceneFrame> {
+    static std::size_t bytes(const AlignedSceneFrame& v) {
+        return sizeof(AlignedSceneFrame) + sae::bytes::of(v.source)
+             + v.faces.size() * sizeof(DetectedFace)
+             + sae::bytes::of(v.crops);
+    }
+};
+
+template<> struct kpn::ChannelDataSize<EmbeddedSceneFrame> {
+    static std::size_t bytes(const EmbeddedSceneFrame& v) {
+        return sizeof(EmbeddedSceneFrame) + sae::bytes::of(v.source)
+             + v.faces.size() * sizeof(DetectedFace)
+             + sae::bytes::of(v.crops)
+             + v.embeddings.size() * sizeof(Embedding);
+    }
+};
+
+template<> struct kpn::ChannelDataSize<TrackedSceneFrame> {
+    static std::size_t bytes(const TrackedSceneFrame& v) {
+        return sizeof(TrackedSceneFrame) + sae::bytes::of(v.source)
+             + v.faces.size() * sizeof(DetectedFace)
+             + sae::bytes::of(v.crops)
+             + v.track_ids.size() * sizeof(int)
+             + v.embeddings.size() * sizeof(Embedding);
+    }
+};
+
+template<> struct kpn::ChannelDataSize<MatchedSceneFrame> {
+    static std::size_t bytes(const MatchedSceneFrame& v) {
+        return sizeof(MatchedSceneFrame) + sae::bytes::of(v.source)
+             + sae::bytes::of(v.actors);
+    }
+};
+
+template<> struct kpn::ChannelDataSize<SceneAnnotation> {
+    static std::size_t bytes(const SceneAnnotation& v) {
+        return sizeof(SceneAnnotation) + sae::bytes::of(v.visible_actors);
+    }
+};
+
+// CutEvent owns nothing on the heap, so the default sizeof(T) is already right.

tests/CMakeLists.txt

@@ -28,6 +28,7 @@ add_executable(sae_tests
     test_embedding_dump.cpp
     test_audio_signature.cpp
     test_benchmark.cpp
+    test_channel_bytes.cpp
     ${CMAKE_SOURCE_DIR}/src/backends/gemm_backend.cpp
     ${CMAKE_SOURCE_DIR}/src/gallery/gallery_store.cpp
     ${CMAKE_SOURCE_DIR}/src/audio_signature.cpp

tests/test_channel_bytes.cpp

@@ -0,0 +1,109 @@
+// Channel byte accounting for the pipeline message types.
+//
+// TRACES: AR-004 | SR-002
+//
+// kpn::ChannelDataSize<T> is what a channel reports as bytes pushed, and its
+// primary template returns sizeof(T). Every message type here is a handful of
+// vectors and a cv::Mat header owning megabytes on the heap, so unspecialised
+// the diagnostics reported ~200 bytes for a message carrying a full decoded
+// frame — off by four orders of magnitude at 1080p.
+//
+// That is the instrument for choosing channel capacities against a memory
+// ceiling, which is the open half of AR-004. These cases assert it measures the
+// payload rather than the header, because a stat that is quietly wrong is worse
+// than no stat: it was read as evidence.
+#include <catch2/catch_test_macros.hpp>
+
+#include <kpn/channel.hpp>
+
+#include "types.hpp"
+
+namespace {
+
+Frame frame_with_image(int w, int h) {
+    Frame f;
+    f.image = cv::Mat(h, w, CV_8UC3, cv::Scalar(0, 0, 0));
+    f.timestamp_sec = 1.0;
+    return f;
+}
+
+}  // namespace
+
+TEST_CASE("frame bytes count the decoded image, not the header", "[channel_bytes]") {
+    const Frame f = frame_with_image(1920, 1080);
+    const std::size_t got = kpn::ChannelDataSize<Frame>::bytes(f);
+
+    // 1920 * 1080 * 3 = 6,220,800 payload bytes.
+    REQUIRE(got >= 1920u * 1080u * 3u);
+    // The header is a rounding error next to it; this is the assertion that
+    // fails on the unspecialised default.
+    CHECK(got > 100u * sizeof(Frame));
+}
+
+TEST_CASE("an empty frame costs only its header", "[channel_bytes]") {
+    // The eof sentinel carries no image, and must not be charged for one.
+    Frame eof;
+    eof.eof = true;
+    CHECK(kpn::ChannelDataSize<Frame>::bytes(eof) == sizeof(Frame));
+}
+
+TEST_CASE("crops and embeddings are counted on top of the frame", "[channel_bytes]") {
+    // The case AR-003 created: a crowd frame occupies one slot exactly as an
+    // empty one does, and only the byte figure distinguishes them.
+    EmbeddedSceneFrame v;
+    v.source = frame_with_image(640, 360);
+    const std::size_t bare = kpn::ChannelDataSize<EmbeddedSceneFrame>::bytes(v);
+
+    constexpr int kFaces = 60;
+    for (int i = 0; i < kFaces; ++i) {
+        v.faces.push_back({});
+        v.crops.emplace_back(112, 112, CV_8UC3, cv::Scalar(0, 0, 0));
+        v.embeddings.emplace_back();
+    }
+    const std::size_t crowded = kpn::ChannelDataSize<EmbeddedSceneFrame>::bytes(v);
+
+    // 60 crops at 112*112*3 = 2,257,920 bytes, plus 60 * 2 KiB of embeddings.
+    CHECK(crowded - bare >= kFaces * (112u * 112u * 3u + sizeof(Embedding)));
+    // And the crowd frame really is the multiple of the empty one that the
+    // item-count capacity cannot see: 640x360x3 is ~691 KB, the crops ~2.26 MB.
+    CHECK(crowded > 3 * bare);
+}
+
+TEST_CASE("every message type on a channel measures its payload", "[channel_bytes]") {
+    // A specialisation missing for any one of these silently reverts that
+    // channel to sizeof(T), which is exactly how this went unnoticed.
+    const Frame f = frame_with_image(320, 240);
+    const std::size_t img = 320u * 240u * 3u;
+
+    SceneFrame sf;        sf.source = f;
+    AlignedSceneFrame af; af.source = f;
+    EmbeddedSceneFrame ef; ef.source = f;
+    TrackedSceneFrame tf; tf.source = f;
+    MatchedSceneFrame mf; mf.source = f;
+
+    CHECK(kpn::ChannelDataSize<SceneFrame>::bytes(sf)         >= img);
+    CHECK(kpn::ChannelDataSize<AlignedSceneFrame>::bytes(af)  >= img);
+    CHECK(kpn::ChannelDataSize<EmbeddedSceneFrame>::bytes(ef) >= img);
+    CHECK(kpn::ChannelDataSize<TrackedSceneFrame>::bytes(tf)  >= img);
+    CHECK(kpn::ChannelDataSize<MatchedSceneFrame>::bytes(mf)  >= img);
+
+    // SceneAnnotation carries no source frame — only the actors it identified,
+    // each with its own crop.
+    SceneAnnotation sa;
+    sa.visible_actors.push_back({});
+    sa.visible_actors.back().crop = cv::Mat(112, 112, CV_8UC3, cv::Scalar(0, 0, 0));
+    CHECK(kpn::ChannelDataSize<SceneAnnotation>::bytes(sa) >= 112u * 112u * 3u);
+}
+
+TEST_CASE("a shared image is charged to each message holding it", "[channel_bytes]") {
+    // cv::Mat is reference-counted, so a frame referenced from several messages
+    // is counted once per reference. The sum is an upper bound on distinct
+    // bytes, and the right bound for "what would this channel keep alive if
+    // nothing else held it" — which is the question a capacity answers.
+    const Frame f = frame_with_image(320, 240);
+    SceneFrame a; a.source = f;
+    SceneFrame b; b.source = f;   // shares the same pixel buffer
+
+    CHECK(kpn::ChannelDataSize<SceneFrame>::bytes(a)
+          == kpn::ChannelDataSize<SceneFrame>::bytes(b));
+}

tests/test_scene_detector_node.cpp

@@ -14,6 +14,9 @@
 
 #include "nodes/scene_detector_node.hpp"
 
+#include <opencv2/imgproc.hpp>
+#include <utility>
+
 #include <vector>
 
 namespace {
@@ -84,3 +87,105 @@ TEST_CASE("too few frames to have a cadence yields an inert window",
 TEST_CASE("a single observed interval is enough", "[scene][AR-011]") {
     CHECK(SceneDetectorFunc::dedup_window_sec({1.0 / 24.0}) == 0.5 / 24.0);
 }
+
+// ── AR-004 — the window stores the model's input, not the decoded frame ───────
+//
+// TRACES: AR-004, AR-010 | SR-002 | UT-003
+//
+// The rolling window held frames as decoded, at full resolution, and left the
+// downscale to the backend — ~590 MB at 1080p to feed a model whose input is
+// 48x27, about 380 KB. Not a channel capacity, so no amount of tuning channel
+// depths would have found it.
+//
+// The risk in fixing it is the project invariant: every model gets the input it
+// was trained for. A model run off-distribution returns confident, plausible,
+// wrong output, and here that means fabricated shot boundaries — which would be
+// indistinguishable from a real cut in the output.
+//
+// So these cases do not check that the frames got smaller. They check that the
+// pixels are *identical* to what the backend would have produced from the full
+// frame, by performing the backend's own two operations independently and
+// comparing byte for byte. Both ort_backend.cpp and trt_backend.cpp guard
+// mis-sized input with convertTo(CV_8UC3) then
+// cv::resize(..., {kFrameW, kFrameH}, 0, 0, cv::INTER_AREA), in that order.
+namespace {
+
+cv::Mat gradient(int w, int h) {
+    // Structured content, not a flat fill: INTER_AREA averages, so a constant
+    // image would compare equal under almost any resize and prove nothing.
+    cv::Mat m(h, w, CV_8UC3);
+    for (int y = 0; y < h; ++y)
+        for (int x = 0; x < w; ++x)
+            m.at<cv::Vec3b>(y, x) = cv::Vec3b(
+                static_cast<uchar>((x * 7 + y * 3) % 256),
+                static_cast<uchar>((x * 13 + y * 5) % 256),
+                static_cast<uchar>((x * 3 + y * 11) % 256));
+    return m;
+}
+
+bool identical(const cv::Mat& a, const cv::Mat& b) {
+    if (a.size() != b.size() || a.type() != b.type()) return false;
+    cv::Mat diff;
+    cv::absdiff(a, b, diff);
+    return cv::countNonZero(diff.reshape(1)) == 0;
+}
+
+}  // namespace
+
+TEST_CASE("the window frame is what the backend would have produced",
+          "[scene][AR-004]") {
+    for (auto [w, h] : {std::pair{1920, 1080}, std::pair{640, 360}, std::pair{720, 480}}) {
+        INFO("source " << w << "x" << h);
+        const cv::Mat full = gradient(w, h);
+
+        // The backend's own guard, performed here independently.
+        cv::Mat expected;
+        cv::resize(full, expected, {ISceneDetector::kFrameW, ISceneDetector::kFrameH},
+                   0, 0, cv::INTER_AREA);
+
+        const cv::Mat got = SceneDetectorFunc::to_model_input(full);
+
+        REQUIRE(got.cols == ISceneDetector::kFrameW);
+        REQUIRE(got.rows == ISceneDetector::kFrameH);
+        REQUIRE(got.type() == CV_8UC3);
+        CHECK(identical(got, expected));
+    }
+}
+
+TEST_CASE("a frame already at model size is passed through untouched",
+          "[scene][AR-004]") {
+    // The backend skips its guard for a correctly-sized frame, so this path must
+    // not resize either — resampling an already-48x27 image would change it.
+    const cv::Mat exact = gradient(ISceneDetector::kFrameW, ISceneDetector::kFrameH);
+    CHECK(identical(SceneDetectorFunc::to_model_input(exact), exact));
+}
+
+TEST_CASE("conversion happens before the resize, as the backend does it",
+          "[scene][AR-004]") {
+    // Order matters: converting a 4-channel frame after downscaling averages
+    // alpha into the colour channels and gives different pixels. The backends
+    // convert first, so this must too.
+    cv::Mat four(360, 640, CV_8UC4, cv::Scalar(10, 20, 30, 255));
+    cv::Mat typed;
+    four.convertTo(typed, CV_8UC3);
+    cv::Mat expected;
+    cv::resize(typed, expected, {ISceneDetector::kFrameW, ISceneDetector::kFrameH},
+               0, 0, cv::INTER_AREA);
+
+    CHECK(identical(SceneDetectorFunc::to_model_input(four), expected));
+}
+
+TEST_CASE("the window's memory is bounded by the model input, not the source",
+          "[scene][AR-004]") {
+    // The point of the change, stated as a number: a full window of 1080p
+    // frames is ~590 MB as decoded and ~380 KB as model input.
+    const cv::Mat full = gradient(1920, 1080);
+    const cv::Mat small = SceneDetectorFunc::to_model_input(full);
+
+    const std::size_t decoded = full.total()  * full.elemSize();
+    const std::size_t stored  = small.total() * small.elemSize();
+
+    INFO("decoded " << decoded << " B, stored " << stored << " B");
+    CHECK(stored * 1000 < decoded);          // three orders of magnitude
+    CHECK(stored == ISceneDetector::kFrameW * ISceneDetector::kFrameH * 3u);
+}

tests/test_track_registry.cpp

@@ -370,3 +370,150 @@ TEST_CASE("confidence grows across frames of the same face", "[registry][AR-025]
     // ...but it must still be worth far less than 50 independent looks would be.
     CHECK(sink.claims[0].effective_obs < 25.0f);
 }
+
+// ── The evidence watermark: presence must not depend on node speed ───────────
+
+/// TRACES: UT-001 | AR-012, AR-013, AR-025 | SR-002
+TEST_CASE("a track is not reaped until the evidence clock passes it",
+          "[registry][AR-013]") {
+    // The tracker and the matcher are separate KPN nodes, and backpressure --
+    // working exactly as AR-004 intends -- lets the tracker run a whole
+    // channel's depth ahead. Reaping on the tracker's clock therefore closed
+    // tracks before their votes arrived: the votes landed on ids that no longer
+    // existed and the run silently under-reported. On the SuperHero fixture that
+    // was 5 actors at channel depth 32 against 0 actors at depth 10322, from
+    // identical input.
+    TrackRegistry reg(cfg(/*extinction=*/5.0), disc());
+    Sink sink; sink.attach(reg);
+    reg.expect_evidence();   // as IdentityMatcherFunc::set_registry does
+
+    int id;
+    {
+        auto s = reg.begin_frame(0.0);
+        id = s.create(0.0, axis(1));
+        s.mark_lost(id, 0.0);
+    }
+
+    // The tracker races 100 s ahead. Nothing has voted yet, so nothing may die:
+    // an unvoted track is not a finished track, it is an unanswered question.
+    { auto s = reg.begin_frame(100.0); (void)s; }
+    CHECK(sink.claims.empty());
+
+    // A vote arriving very late still lands, because the track is still there.
+    reg.observe(id, /*actor*/ 3, /*posterior*/ 0.99f, axis(1));
+    CHECK(reg.dropped_votes() == 0);
+
+    // Only once the evidence clock passes last_seen + extinction does it close.
+    reg.advance_evidence(4.0);
+    CHECK(sink.claims.empty());
+    reg.advance_evidence(6.0);
+    REQUIRE(sink.claims.size() == 1);
+    CHECK(sink.claims[0].actor_idx == 3);
+    // AR-013 still holds: the window ends at the last sighting, never at the
+    // moment of death, and never at the watermark that authorised it.
+    CHECK(sink.claims[0].last_seen == 0.0);
+}
+
+/// TRACES: UT-001 | AR-008, AR-013 | SR-002
+TEST_CASE("a track retired from association is still open to evidence",
+          "[registry][AR-008]") {
+    // The two clocks answer different questions and must not share an answer.
+    // Association asks "may this detection link to that track?" on the tracker's
+    // clock; reaping asks "is that track finished?" and cannot answer until the
+    // votes are in. Deferring both to the evidence clock was the second half of
+    // this bug: retired tracks lingered in the candidate pool for as long as the
+    // matcher lagged, so a new face re-associated onto a long-dead track and two
+    // people merged into one window.
+    TrackRegistry reg(cfg(/*extinction=*/5.0), disc());
+    Sink sink; sink.attach(reg);
+    reg.expect_evidence();
+
+    int id;
+    {
+        auto s = reg.begin_frame(0.0);
+        id = s.create(0.0, axis(1));
+        s.mark_lost(id, 0.0);
+    }
+
+    {
+        auto s = reg.begin_frame(3.0);          // inside the window
+        CHECK(s.candidates().size() == 1);      // still associable
+    }
+    {
+        auto s = reg.begin_frame(50.0);         // far outside it
+        CHECK(s.candidates().empty());          // retired from association...
+    }
+    // ...but not gone, and still able to receive the votes in flight for it.
+    reg.observe(id, 7, 0.99f, axis(1));
+    CHECK(reg.dropped_votes() == 0);
+    reg.advance_evidence(50.0);
+    REQUIRE(sink.claims.size() == 1);
+    CHECK(sink.claims[0].actor_idx == 7);
+}
+
+// ── AR-025 — non-matches must not spend an actor's evidence budget ───────────
+TEST_CASE("frames that recognise nobody do not exhaust the budget",
+          "[registry][AR-025]") {
+    // The correlation discount is an effective-sample correction: with
+    // observations correlated at rho, the n-th is worth
+    // n_eff(n+1) - n_eff(n) = 2/((n+1)(n+2)) at rho=0.5, so it decays
+    // quadratically and the total converges to 1/rho = 2. That saturation is
+    // deliberate — a long static shot must not out-argue varied evidence by
+    // lasting longer.
+    //
+    // What was not deliberate is that every scored face spent it, including
+    // ones that matched nobody. An observation at p=0.02 contributes
+    // log(0.98) = -0.02 of belief — nothing — while consuming the same
+    // increment as one at p=0.95. Measured on SuperHero-2: 103 observations on
+    // one track, effective weight 2.026, belief 0.455 against a 0.881
+    // threshold, with the 51 frames that *did* identify the actor arriving
+    // when each was worth 0.0002. The identification was lost.
+    //
+    // It also made the answer depend on frame rate — deliver more frames,
+    // dilute the budget with more non-matches, and a track that was owned stops
+    // being owned — which is the defect AR-013 already had to fix once.
+    TrackRegistry reg(cfg(), disc());
+    Sink sink; sink.attach(reg);
+
+    int id;
+    { auto f = reg.begin_frame(0.0); id = f.create(0.0, axis(0)); }
+
+    // A long run of frames that match nobody: the detector saw a face, the
+    // matcher could not place it. These are not evidence for actor 1.
+    for (int i = 0; i < 40; ++i) reg.observe(id, 1, 0.02f, axis(0));
+
+    // Then the actor is clearly recognised. Before this fix the budget was
+    // already spent and these could not move the belief.
+    for (int i = 0; i < 6; ++i) reg.observe(id, 1, 0.9f, axis(0));
+
+    reg.flush(1.0);
+
+    REQUIRE(sink.claims.size() == 1);
+    INFO("belief " << sink.claims[0].belief
+         << " effective_obs " << sink.claims[0].effective_obs);
+    CHECK(sink.claims[0].actor_idx == 1);
+    CHECK(sink.claims[0].belief > 0.88f);
+}
+
+TEST_CASE("a near-miss is still evidence", "[registry][AR-025]") {
+    // The floor is at 0.5 — where the posterior stops favouring the hypothesis
+    // — not at the matcher's acceptance threshold. A run of near-misses for one
+    // actor is informative and must still accumulate, which is the property the
+    // identity matcher's comment relies on when it feeds every scored face
+    // rather than only the accepted ones.
+    TrackRegistry reg(cfg(), disc());
+    Sink sink; sink.attach(reg);
+
+    int id;
+    { auto f = reg.begin_frame(0.0); id = f.create(0.0, axis(0)); }
+    // 0.7 is below the matcher's acceptance threshold (0.754 on the SuperHero
+    // gallery) and above the 0.5 floor: a frame that would not be reported as
+    // an identification, but is still evidence. Twelve of them accumulate to
+    // ~0.89, past the 0.881 ownership threshold.
+    for (int i = 0; i < 12; ++i) reg.observe(id, 3, 0.7f, axis(0));
+    reg.flush(1.0);
+
+    REQUIRE(sink.claims.size() == 1);
+    INFO("belief " << sink.claims[0].belief);
+    CHECK(sink.claims[0].actor_idx == 3);   // owned on near-misses alone
+}