Files
DarkRoom/core/dr-face/Cargo.toml
T
dtourolleandClaude Opus 5 19981c1033 Detect, align and embed faces with SCRFD and MobileFaceNet
Ports the pipeline from the C++ reference in ../scene-actor-extraction
(MIT, same author). End to end on real portraits it separates identities
the way the reference's fitted calibration says it should: 0.596 between
distinct photographs of one person, 0.05 between different people, either
side of MBF's 0.267 boundary.

Three things are structural rather than incidental:

Aligned112 can only be built by align::warp, so Embedder::embed cannot be
handed an unaligned bounding-box crop. That mistake yields 512 plausible
unit-norm numbers and no error, so the type system refuses it instead.

Embedding carries its ModelId and cosine() returns None across models,
because a cross-model similarity is the one mistake that produces
plausible garbage rather than a failure.

The model-free half -- alignment, embedding arithmetic, f16 storage --
sits outside the inference feature and is covered by 11 tests that need
no weights on the machine.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-26 19:57:56 +02:00

48 lines
1.6 KiB
TOML

[package]
name = "dr-face"
version.workspace = true
edition.workspace = true
rust-version.workspace = true
license.workspace = true
[dependencies]
thiserror.workspace = true
log.workspace = true
# Inference. `ort` is the API; **tract is the engine** — see the workspace
# manifest, and docs/faces.md §3, for why the C++ ONNX Runtime is not linked.
ort = { workspace = true, optional = true }
ort-tract = { workspace = true, optional = true }
ndarray = { workspace = true, optional = true }
[dev-dependencies]
zune-jpeg.workspace = true
env_logger.workspace = true
# The M1 probe drives `ort` directly so it can print the raw load error.
ort = { workspace = true }
ort-tract = { workspace = true }
[[example]]
name = "probe"
required-features = ["inference"]
[[example]]
name = "faces"
required-features = ["inference"]
[features]
# Nothing on by default, and in particular **no `embedded-model`**: the weights
# are not a build input and never become one (docs/faces.md §2.2). A feature
# flag that *could* embed them is a flag someone eventually sets in a packaging
# script, and the InsightFace grant does not survive that.
default = []
# The ONNX runtime, and the two stages that need it.
#
# Separable because the accuracy of this subsystem lives in `calibrate` and
# `cluster`, which are arithmetic over embeddings with no model in them. They
# must be testable against synthetic embeddings on a machine with no weights on
# it — a test suite that needs a research-licensed download is a test suite
# that does not run in CI.
inference = ["dep:ort", "dep:ort-tract", "dep:ndarray"]