1897 lines
70 KiB
HTML
1897 lines
70 KiB
HTML
|
||
<!doctype html>
|
||
<html lang="en" class="no-js">
|
||
<head>
|
||
|
||
<meta charset="utf-8">
|
||
<meta name="viewport" content="width=device-width,initial-scale=1">
|
||
|
||
<meta name="description" content="Face-recognition pipeline for finding on-screen actor presence in film/TV, built on KPN++">
|
||
|
||
|
||
|
||
<link rel="canonical" href="https://pages.tourolle.paris/dtourolle/scene-actor-extraction/model-bakeoff/">
|
||
|
||
|
||
<link rel="prev" href="../lvface-deep-dive/">
|
||
|
||
|
||
<link rel="next" href="../optimizer-experiments/">
|
||
|
||
|
||
|
||
|
||
|
||
<link rel="icon" href="../assets/images/favicon.png">
|
||
<meta name="generator" content="mkdocs-1.6.1, mkdocs-material-9.7.7">
|
||
|
||
|
||
|
||
<title>Model Bake-off & Re-tune (full log) - scene-actor-extraction</title>
|
||
|
||
|
||
|
||
<link rel="stylesheet" href="../assets/stylesheets/main.ec1eaa64.min.css">
|
||
|
||
|
||
<link rel="stylesheet" href="../assets/stylesheets/palette.ab4e12ef.min.css">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<link rel="preconnect" href="https://fonts.gstatic.com" crossorigin>
|
||
<link rel="stylesheet" href="https://fonts.googleapis.com/css?family=Roboto:300,300i,400,400i,700,700i%7CRoboto+Mono:400,400i,700,700i&display=fallback">
|
||
<style>:root{--md-text-font:"Roboto";--md-code-font:"Roboto Mono"}</style>
|
||
|
||
|
||
|
||
<link rel="stylesheet" href="../stylesheets/extra.css">
|
||
|
||
<script>__md_scope=new URL("..",location),__md_hash=e=>[...e].reduce(((e,_)=>(e<<5)-e+_.charCodeAt(0)),0),__md_get=(e,_=localStorage,t=__md_scope)=>JSON.parse(_.getItem(t.pathname+"."+e)),__md_set=(e,_,t=localStorage,a=__md_scope)=>{try{t.setItem(a.pathname+"."+e,JSON.stringify(_))}catch(e){}}</script>
|
||
|
||
|
||
|
||
|
||
|
||
</head>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<body dir="ltr" data-md-color-scheme="slate" data-md-color-primary="black" data-md-color-accent="amber">
|
||
|
||
|
||
<input class="md-toggle" data-md-toggle="drawer" type="checkbox" id="__drawer" autocomplete="off">
|
||
<input class="md-toggle" data-md-toggle="search" type="checkbox" id="__search" autocomplete="off">
|
||
<label class="md-overlay" for="__drawer"></label>
|
||
<div data-md-component="skip">
|
||
|
||
|
||
<a href="#model-bake-off-threshold-re-tune-experiment-log-2026-07-1819" class="md-skip">
|
||
Skip to content
|
||
</a>
|
||
|
||
</div>
|
||
<div data-md-component="announce">
|
||
|
||
</div>
|
||
|
||
|
||
|
||
|
||
<header class="md-header" data-md-component="header">
|
||
<nav class="md-header__inner md-grid" aria-label="Header">
|
||
<a href=".." title="scene-actor-extraction" class="md-header__button md-logo" aria-label="scene-actor-extraction" data-md-component="logo">
|
||
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M12 8a3 3 0 0 0 3-3 3 3 0 0 0-3-3 3 3 0 0 0-3 3 3 3 0 0 0 3 3m0 3.54C9.64 9.35 6.5 8 3 8v11c3.5 0 6.64 1.35 9 3.54 2.36-2.19 5.5-3.54 9-3.54V8c-3.5 0-6.64 1.35-9 3.54"/></svg>
|
||
|
||
</a>
|
||
<label class="md-header__button md-icon" for="__drawer">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M3 6h18v2H3zm0 5h18v2H3zm0 5h18v2H3z"/></svg>
|
||
</label>
|
||
<div class="md-header__title" data-md-component="header-title">
|
||
<div class="md-header__ellipsis">
|
||
<div class="md-header__topic">
|
||
<span class="md-ellipsis">
|
||
scene-actor-extraction
|
||
</span>
|
||
</div>
|
||
<div class="md-header__topic" data-md-component="header-topic">
|
||
<span class="md-ellipsis">
|
||
|
||
Model Bake-off & Re-tune (full log)
|
||
|
||
</span>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
|
||
|
||
<form class="md-header__option" data-md-component="palette">
|
||
|
||
|
||
|
||
|
||
<input class="md-option" data-md-color-media="(prefers-color-scheme: dark)" data-md-color-scheme="slate" data-md-color-primary="black" data-md-color-accent="amber" aria-label="Switch to light mode" type="radio" name="__palette" id="__palette_0">
|
||
|
||
<label class="md-header__button md-icon" title="Switch to light mode" for="__palette_1" hidden>
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M12 7a5 5 0 0 1 5 5 5 5 0 0 1-5 5 5 5 0 0 1-5-5 5 5 0 0 1 5-5m0 2a3 3 0 0 0-3 3 3 3 0 0 0 3 3 3 3 0 0 0 3-3 3 3 0 0 0-3-3m0-7 2.39 3.42C13.65 5.15 12.84 5 12 5s-1.65.15-2.39.42zM3.34 7l4.16-.35A7.2 7.2 0 0 0 5.94 8.5c-.44.74-.69 1.5-.83 2.29zm.02 10 1.76-3.77a7.131 7.131 0 0 0 2.38 4.14zM20.65 7l-1.77 3.79a7.02 7.02 0 0 0-2.38-4.15zm-.01 10-4.14.36c.59-.51 1.12-1.14 1.54-1.86.42-.73.69-1.5.83-2.29zM12 22l-2.41-3.44c.74.27 1.55.44 2.41.44.82 0 1.63-.17 2.37-.44z"/></svg>
|
||
</label>
|
||
|
||
|
||
|
||
|
||
|
||
<input class="md-option" data-md-color-media="(prefers-color-scheme: light)" data-md-color-scheme="default" data-md-color-primary="black" data-md-color-accent="indigo" aria-label="Switch to dark mode" type="radio" name="__palette" id="__palette_1">
|
||
|
||
<label class="md-header__button md-icon" title="Switch to dark mode" for="__palette_0" hidden>
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="m17.75 4.09-2.53 1.94.91 3.06-2.63-1.81-2.63 1.81.91-3.06-2.53-1.94L12.44 4l1.06-3 1.06 3zm3.5 6.91-1.64 1.25.59 1.98-1.7-1.17-1.7 1.17.59-1.98L15.75 11l2.06-.05L18.5 9l.69 1.95zm-2.28 4.95c.83-.08 1.72 1.1 1.19 1.85-.32.45-.66.87-1.08 1.27C15.17 23 8.84 23 4.94 19.07c-3.91-3.9-3.91-10.24 0-14.14.4-.4.82-.76 1.27-1.08.75-.53 1.93.36 1.85 1.19-.27 2.86.69 5.83 2.89 8.02a9.96 9.96 0 0 0 8.02 2.89m-1.64 2.02a12.08 12.08 0 0 1-7.8-3.47c-2.17-2.19-3.33-5-3.49-7.82-2.81 3.14-2.7 7.96.31 10.98 3.02 3.01 7.84 3.12 10.98.31"/></svg>
|
||
</label>
|
||
|
||
|
||
</form>
|
||
|
||
|
||
|
||
<script>var palette=__md_get("__palette");if(palette&&palette.color){if("(prefers-color-scheme)"===palette.color.media){var media=matchMedia("(prefers-color-scheme: light)"),input=document.querySelector(media.matches?"[data-md-color-media='(prefers-color-scheme: light)']":"[data-md-color-media='(prefers-color-scheme: dark)']");palette.color.media=input.getAttribute("data-md-color-media"),palette.color.scheme=input.getAttribute("data-md-color-scheme"),palette.color.primary=input.getAttribute("data-md-color-primary"),palette.color.accent=input.getAttribute("data-md-color-accent")}for(var[key,value]of Object.entries(palette.color))document.body.setAttribute("data-md-color-"+key,value)}</script>
|
||
|
||
|
||
|
||
|
||
|
||
<label class="md-header__button md-icon" for="__search">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M9.5 3A6.5 6.5 0 0 1 16 9.5c0 1.61-.59 3.09-1.56 4.23l.27.27h.79l5 5-1.5 1.5-5-5v-.79l-.27-.27A6.52 6.52 0 0 1 9.5 16 6.5 6.5 0 0 1 3 9.5 6.5 6.5 0 0 1 9.5 3m0 2C7 5 5 7 5 9.5S7 14 9.5 14 14 12 14 9.5 12 5 9.5 5"/></svg>
|
||
</label>
|
||
<div class="md-search" data-md-component="search" role="dialog">
|
||
<label class="md-search__overlay" for="__search"></label>
|
||
<div class="md-search__inner" role="search">
|
||
<form class="md-search__form" name="search">
|
||
<input type="text" class="md-search__input" name="query" aria-label="Search" placeholder="Search" autocapitalize="off" autocorrect="off" autocomplete="off" spellcheck="false" data-md-component="search-query" required>
|
||
<label class="md-search__icon md-icon" for="__search">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M9.5 3A6.5 6.5 0 0 1 16 9.5c0 1.61-.59 3.09-1.56 4.23l.27.27h.79l5 5-1.5 1.5-5-5v-.79l-.27-.27A6.52 6.52 0 0 1 9.5 16 6.5 6.5 0 0 1 3 9.5 6.5 6.5 0 0 1 9.5 3m0 2C7 5 5 7 5 9.5S7 14 9.5 14 14 12 14 9.5 12 5 9.5 5"/></svg>
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M20 11v2H8l5.5 5.5-1.42 1.42L4.16 12l7.92-7.92L13.5 5.5 8 11z"/></svg>
|
||
</label>
|
||
<nav class="md-search__options" aria-label="Search">
|
||
|
||
<button type="reset" class="md-search__icon md-icon" title="Clear" aria-label="Clear" tabindex="-1">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M19 6.41 17.59 5 12 10.59 6.41 5 5 6.41 10.59 12 5 17.59 6.41 19 12 13.41 17.59 19 19 17.59 13.41 12z"/></svg>
|
||
</button>
|
||
</nav>
|
||
|
||
</form>
|
||
<div class="md-search__output">
|
||
<div class="md-search__scrollwrap" tabindex="0" data-md-scrollfix>
|
||
<div class="md-search-result" data-md-component="search-result">
|
||
<div class="md-search-result__meta">
|
||
Initializing search
|
||
</div>
|
||
<ol class="md-search-result__list" role="presentation"></ol>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
|
||
|
||
|
||
<div class="md-header__source">
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction" title="Go to repository" class="md-source" data-md-component="source">
|
||
<div class="md-source__icon md-icon">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 448 512"><!--! Font Awesome Free 7.1.0 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free (Icons: CC BY 4.0, Fonts: SIL OFL 1.1, Code: MIT License) Copyright 2025 Fonticons, Inc.--><path d="M439.6 236.1 244 40.5c-5.4-5.5-12.8-8.5-20.4-8.5s-15 3-20.4 8.4L162.5 81l51.5 51.5c27.1-9.1 52.7 16.8 43.4 43.7l49.7 49.7c34.2-11.8 61.2 31 35.5 56.7-26.5 26.5-70.2-2.9-56-37.3L240.3 199v121.9c25.3 12.5 22.3 41.8 9.1 55-6.4 6.4-15.2 10.1-24.3 10.1s-17.8-3.6-24.3-10.1c-17.6-17.6-11.1-46.9 11.2-56v-123c-20.8-8.5-24.6-30.7-18.6-45L142.6 101 8.5 235.1C3 240.6 0 247.9 0 255.5s3 15 8.5 20.4l195.6 195.7c5.4 5.4 12.7 8.4 20.4 8.4s15-3 20.4-8.4l194.7-194.7c5.4-5.4 8.4-12.8 8.4-20.4s-3-15-8.4-20.4"/></svg>
|
||
</div>
|
||
<div class="md-source__repository">
|
||
dtourolle/scene-actor-extraction
|
||
</div>
|
||
</a>
|
||
</div>
|
||
|
||
</nav>
|
||
|
||
</header>
|
||
|
||
<div class="md-container" data-md-component="container">
|
||
|
||
|
||
|
||
|
||
|
||
<nav class="md-tabs" aria-label="Tabs" data-md-component="tabs">
|
||
<div class="md-grid">
|
||
<ul class="md-tabs__list">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-tabs__item">
|
||
<a href=".." class="md-tabs__link">
|
||
|
||
|
||
|
||
|
||
|
||
Home
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-tabs__item">
|
||
<a href="../best-model/" class="md-tabs__link">
|
||
|
||
|
||
|
||
Findings
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-tabs__item md-tabs__item--active">
|
||
<a href="./" class="md-tabs__link">
|
||
|
||
|
||
|
||
|
||
|
||
Model Bake-off & Re-tune (full log)
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-tabs__item">
|
||
<a href="../optimizer-experiments/" class="md-tabs__link">
|
||
|
||
|
||
|
||
|
||
|
||
Optimizer Experiments (prior round)
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-tabs__item">
|
||
<a href="../service-conversion/" class="md-tabs__link">
|
||
|
||
|
||
|
||
|
||
|
||
Service Conversion (proposal)
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
</ul>
|
||
</div>
|
||
</nav>
|
||
|
||
|
||
|
||
<main class="md-main" data-md-component="main">
|
||
<div class="md-main__inner md-grid">
|
||
|
||
|
||
|
||
<div class="md-sidebar md-sidebar--primary" data-md-component="sidebar" data-md-type="navigation" >
|
||
<div class="md-sidebar__scrollwrap">
|
||
<div class="md-sidebar__inner">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<nav class="md-nav md-nav--primary md-nav--lifted" aria-label="Navigation" data-md-level="0">
|
||
<label class="md-nav__title" for="__drawer">
|
||
<a href=".." title="scene-actor-extraction" class="md-nav__button md-logo" aria-label="scene-actor-extraction" data-md-component="logo">
|
||
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M12 8a3 3 0 0 0 3-3 3 3 0 0 0-3-3 3 3 0 0 0-3 3 3 3 0 0 0 3 3m0 3.54C9.64 9.35 6.5 8 3 8v11c3.5 0 6.64 1.35 9 3.54 2.36-2.19 5.5-3.54 9-3.54V8c-3.5 0-6.64 1.35-9 3.54"/></svg>
|
||
|
||
</a>
|
||
scene-actor-extraction
|
||
</label>
|
||
|
||
<div class="md-nav__source">
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction" title="Go to repository" class="md-source" data-md-component="source">
|
||
<div class="md-source__icon md-icon">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 448 512"><!--! Font Awesome Free 7.1.0 by @fontawesome - https://fontawesome.com License - https://fontawesome.com/license/free (Icons: CC BY 4.0, Fonts: SIL OFL 1.1, Code: MIT License) Copyright 2025 Fonticons, Inc.--><path d="M439.6 236.1 244 40.5c-5.4-5.5-12.8-8.5-20.4-8.5s-15 3-20.4 8.4L162.5 81l51.5 51.5c27.1-9.1 52.7 16.8 43.4 43.7l49.7 49.7c34.2-11.8 61.2 31 35.5 56.7-26.5 26.5-70.2-2.9-56-37.3L240.3 199v121.9c25.3 12.5 22.3 41.8 9.1 55-6.4 6.4-15.2 10.1-24.3 10.1s-17.8-3.6-24.3-10.1c-17.6-17.6-11.1-46.9 11.2-56v-123c-20.8-8.5-24.6-30.7-18.6-45L142.6 101 8.5 235.1C3 240.6 0 247.9 0 255.5s3 15 8.5 20.4l195.6 195.7c5.4 5.4 12.7 8.4 20.4 8.4s15-3 20.4-8.4l194.7-194.7c5.4-5.4 8.4-12.8 8.4-20.4s-3-15-8.4-20.4"/></svg>
|
||
</div>
|
||
<div class="md-source__repository">
|
||
dtourolle/scene-actor-extraction
|
||
</div>
|
||
</a>
|
||
</div>
|
||
|
||
<ul class="md-nav__list" data-md-scrollfix>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href=".." class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Home
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item md-nav__item--nested">
|
||
|
||
|
||
|
||
<input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_2" >
|
||
|
||
|
||
<label class="md-nav__link" for="__nav_2" id="__nav_2_label" tabindex="0">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Findings
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
<span class="md-nav__icon md-icon"></span>
|
||
</label>
|
||
|
||
<nav class="md-nav" data-md-level="1" aria-labelledby="__nav_2_label" aria-expanded="false">
|
||
<label class="md-nav__title" for="__nav_2">
|
||
<span class="md-nav__icon md-icon"></span>
|
||
|
||
|
||
Findings
|
||
|
||
|
||
</label>
|
||
<ul class="md-nav__list" data-md-scrollfix>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../best-model/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Best Model
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../gallery-scope/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Gallery Scope (Full vs. Limited)
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../pose-expansion/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Pose Expansion
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../lvface-deep-dive/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
LVFace Deep Dive
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
</ul>
|
||
</nav>
|
||
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item md-nav__item--active">
|
||
|
||
<input class="md-nav__toggle md-toggle" type="checkbox" id="__toc">
|
||
|
||
|
||
|
||
|
||
|
||
<label class="md-nav__link md-nav__link--active" for="__toc">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Model Bake-off & Re-tune (full log)
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
<span class="md-nav__icon md-icon"></span>
|
||
</label>
|
||
|
||
<a href="./" class="md-nav__link md-nav__link--active">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Model Bake-off & Re-tune (full log)
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
|
||
|
||
|
||
<nav class="md-nav md-nav--secondary" aria-label="Table of contents">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<label class="md-nav__title" for="__toc">
|
||
<span class="md-nav__icon md-icon"></span>
|
||
Table of contents
|
||
</label>
|
||
<ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#why-this-experiment-and-what-it-actually-delivered" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Why this experiment, and what it actually delivered
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#tldr-what-changed-in-srcconfighpp" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
TL;DR — what changed in src/config.hpp
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#why-re-run-at-all" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Why re-run at all
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#the-deadlock-that-was-blocking-all-of-this" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
The deadlock that was blocking all of this
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#concurrency-tuning" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Concurrency tuning
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#training-films-and-validation-set" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Training films and validation set
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#search-space-and-de-settings" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Search space and DE settings
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#results-all-16-combos-4-models-full-restricted-expand-noexp" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Results — all 16 combos (4 models × {full, restricted} × {expand, noexp})
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#calibration-curves-discriminative-power-independent-of-the-threshold" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Calibration curves — discriminative power, independent of the threshold
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#two-effects-in-isolation-gallery-scope-and-pose-expansion" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Two effects in isolation: gallery scope, and pose expansion
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#what-the-data-says" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
What the data says
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#held-out-validation-the-number-that-actually-matters" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Held-out validation — the number that actually matters
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#caveats-what-this-is-not" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Caveats / what this is not
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#reproduce" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Reproduce
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
</ul>
|
||
|
||
</nav>
|
||
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../optimizer-experiments/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Optimizer Experiments (prior round)
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<li class="md-nav__item">
|
||
<a href="../service-conversion/" class="md-nav__link">
|
||
|
||
|
||
|
||
<span class="md-ellipsis">
|
||
|
||
|
||
Service Conversion (proposal)
|
||
|
||
|
||
|
||
</span>
|
||
|
||
|
||
|
||
</a>
|
||
</li>
|
||
|
||
|
||
|
||
</ul>
|
||
</nav>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
|
||
|
||
|
||
<div class="md-sidebar md-sidebar--secondary" data-md-component="sidebar" data-md-type="toc" >
|
||
<div class="md-sidebar__scrollwrap">
|
||
<div class="md-sidebar__inner">
|
||
|
||
|
||
<nav class="md-nav md-nav--secondary" aria-label="Table of contents">
|
||
|
||
|
||
|
||
|
||
|
||
|
||
<label class="md-nav__title" for="__toc">
|
||
<span class="md-nav__icon md-icon"></span>
|
||
Table of contents
|
||
</label>
|
||
<ul class="md-nav__list" data-md-component="toc" data-md-scrollfix>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#why-this-experiment-and-what-it-actually-delivered" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Why this experiment, and what it actually delivered
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#tldr-what-changed-in-srcconfighpp" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
TL;DR — what changed in src/config.hpp
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#why-re-run-at-all" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Why re-run at all
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#the-deadlock-that-was-blocking-all-of-this" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
The deadlock that was blocking all of this
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#concurrency-tuning" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Concurrency tuning
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#training-films-and-validation-set" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Training films and validation set
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#search-space-and-de-settings" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Search space and DE settings
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#results-all-16-combos-4-models-full-restricted-expand-noexp" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Results — all 16 combos (4 models × {full, restricted} × {expand, noexp})
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#calibration-curves-discriminative-power-independent-of-the-threshold" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Calibration curves — discriminative power, independent of the threshold
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#two-effects-in-isolation-gallery-scope-and-pose-expansion" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Two effects in isolation: gallery scope, and pose expansion
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#what-the-data-says" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
What the data says
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#held-out-validation-the-number-that-actually-matters" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Held-out validation — the number that actually matters
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#caveats-what-this-is-not" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Caveats / what this is not
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
<li class="md-nav__item">
|
||
<a href="#reproduce" class="md-nav__link">
|
||
<span class="md-ellipsis">
|
||
|
||
Reproduce
|
||
|
||
</span>
|
||
</a>
|
||
|
||
</li>
|
||
|
||
</ul>
|
||
|
||
</nav>
|
||
</div>
|
||
</div>
|
||
</div>
|
||
|
||
|
||
|
||
<div class="md-content" data-md-component="content">
|
||
|
||
<article class="md-content__inner md-typeset">
|
||
|
||
|
||
|
||
|
||
|
||
<h1 id="model-bake-off-threshold-re-tune-experiment-log-2026-07-1819">Model bake-off + threshold re-tune — experiment log (2026-07-18/19)<a class="headerlink" href="#model-bake-off-threshold-re-tune-experiment-log-2026-07-1819" title="Permanent link">¶</a></h1>
|
||
<p>Follow-on to <a href="../optimizer-experiments/">the prior optimizer round</a>, which used
|
||
an older, since-superseded scene-union metric. This round uses the <strong>per-second</strong> metric
|
||
(<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/scripts/optimizer/second_score.py"><code>scripts/optimizer/second_score.py</code></a>)
|
||
and answers three questions in one 16-run matrix: which embedding model is best, does cast-restriction cut misIDs, and does
|
||
per-film gallery expansion help.</p>
|
||
<p>(This was the fourth optimizer campaign against the X-Ray corpus, so its on-disk
|
||
artifacts carry an internal <code>rep4_</code> prefix — <code>experiments/results/rep4_best_*.json</code>,
|
||
<code>experiments/trajectories/rep4_*.jsonl</code>, and the manifests referenced below. The
|
||
earlier campaigns used the superseded scene-union metric and were discarded.)</p>
|
||
<h2 id="why-this-experiment-and-what-it-actually-delivered">Why this experiment, and what it actually delivered<a class="headerlink" href="#why-this-experiment-and-what-it-actually-delivered" title="Permanent link">¶</a></h2>
|
||
<p>Four goals going in, and an honest read on each after held-out validation (see
|
||
below):</p>
|
||
<ol>
|
||
<li><strong>Find the best default parameters to ship.</strong> Partially delivered. The DE optimum
|
||
generalizes <em>unevenly</em> — strong on 3 of 5 held-out films, badly broken on 2 (one
|
||
with a 974-count misID blowup). The tuned values are shipped anyway (see
|
||
Caveats) because they still beat the old defaults on average, but this is not a
|
||
settled, film-agnostic optimum.</li>
|
||
<li><strong>Find the best default model.</strong> Delivered with more confidence. LVFace beat
|
||
r50/r18/mbf across all 4 training combos, and nothing in held-out validation
|
||
contradicts the model choice specifically — the held-out failures trace to
|
||
<code>extinction_sec</code>/threshold interactions and gallery coverage, not the embedder.</li>
|
||
<li><strong>Provide insight into how the application works.</strong> The strongest, most durable
|
||
output. Found and fixed a real teardown deadlock bug (100% reproducible, not the
|
||
assumed rare GPU flake), established a real concurrency ceiling (8 parallel
|
||
replays, not more), and found a real parameter interaction (a strict
|
||
<code>prob_threshold</code> "earns" a longer extinction window before it starts hurting).</li>
|
||
<li><strong>Demonstrate limitations.</strong> Delivered, and reinforced hard by held-out
|
||
validation — see the "Held-out validation" section below for concrete examples,
|
||
including a screenshot of the matcher naming 15 actors, none correctly, on a
|
||
completely blank title card.</li>
|
||
</ol>
|
||
<h2 id="tldr-what-changed-in-srcconfighpp">TL;DR — what changed in <code>src/config.hpp</code><a class="headerlink" href="#tldr-what-changed-in-srcconfighpp" title="Permanent link">¶</a></h2>
|
||
<table>
|
||
<thead>
|
||
<tr>
|
||
<th>knob</th>
|
||
<th>old default</th>
|
||
<th>new default</th>
|
||
<th>why</th>
|
||
</tr>
|
||
</thead>
|
||
<tbody>
|
||
<tr>
|
||
<td><code>arcface_model</code></td>
|
||
<td><code>arcface_w600k_r50.onnx</code></td>
|
||
<td><strong><code>LVFace-B_Glint360K.onnx</code></strong></td>
|
||
<td>Best F1 in the full-gallery bake-off (75.3% vs r50's 68.5%). LVFace was worth its size.</td>
|
||
</tr>
|
||
<tr>
|
||
<td><code>prob_threshold</code></td>
|
||
<td>0.76</td>
|
||
<td><strong>0.754</strong></td>
|
||
<td>Re-tuned for LVFace + per-second metric.</td>
|
||
</tr>
|
||
<tr>
|
||
<td><code>extinction_sec</code></td>
|
||
<td>1.5</td>
|
||
<td><strong>57.4</strong></td>
|
||
<td>Reverses the earlier "short is better" finding — see below.</td>
|
||
</tr>
|
||
<tr>
|
||
<td><code>anneal_sec</code></td>
|
||
<td>10.0</td>
|
||
<td><strong>35.5</strong></td>
|
||
<td>Same reversal; previously thought insensitive.</td>
|
||
</tr>
|
||
<tr>
|
||
<td><code>expand_gallery</code></td>
|
||
<td>false</td>
|
||
<td><strong>true</strong></td>
|
||
<td>Helped recall on the full (unrestricted) gallery for the winning model — opposite of the earlier assumption.</td>
|
||
</tr>
|
||
</tbody>
|
||
</table>
|
||
<p>These are the <strong><code>LVFace-B_Glint360K_full_exp</code></strong> winning values, applied to
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/src/config.hpp"><code>src/config.hpp</code></a> — the best result that
|
||
uses only features already live in the running app (full gallery, no cast
|
||
restriction; see below for why restricted mode isn't applied even though it scored
|
||
higher).</p>
|
||
<h2 id="why-re-run-at-all">Why re-run at all<a class="headerlink" href="#why-re-run-at-all" title="Permanent link">¶</a></h2>
|
||
<p><a href="../optimizer-experiments/">The prior round</a>'s scene-union metric hid out-of-cast false positives
|
||
behind a gallery∩cast recall mask, and the earlier 9-film benchmark was scene-level
|
||
(union over a whole X-Ray scene), not a fair per-timepoint comparison. SESSION_STATE
|
||
flagged the old scene-metric bake-off numbers (R50≈LVFace≈MBF ~85%) as superseded.
|
||
This round uses <code>second_score.py</code>: uniform per-second sampling, GT = X-Ray scene's
|
||
cast at time <em>t</em>, pred = actors whose presence window covers <em>t</em>, FPI weighted 10×
|
||
when the named actor isn't in the film's cast at all (true misID) vs. an in-cast
|
||
timing slip. FN only counts gallery-known cast (fair recall — 67% of X-Ray cast have
|
||
no reference embedding; see
|
||
<a href="../optimizer-experiments/#the-gallery-coverage-gap">the prior round's gallery-coverage-gap analysis</a>).</p>
|
||
<h2 id="the-deadlock-that-was-blocking-all-of-this">The deadlock that was blocking all of this<a class="headerlink" href="#the-deadlock-that-was-blocking-all-of-this" title="Permanent link">¶</a></h2>
|
||
<p>Every replay in this line of work goes through
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/scripts/optimizer/replay.py"><code>scripts/optimizer/replay.py</code></a>,
|
||
which runs the real C++ tracker/matcher/scene_tracker nodes inside a
|
||
Python-assembled KPN network. Before this session, every subprocess replay <strong>timed out at 45s, 100% of
|
||
the time</strong> — not the documented ~20-30% ROCm rocBLAS-GEMM driver flake, but a plain
|
||
logic bug: <code>replay.py</code>'s CLI called <code>replay(net, ..., stop=False)</code> to <em>skip</em>
|
||
<code>net.stop()</code> (trying to dodge the GEMM deadlock), planning to <code>os._exit(0)</code>
|
||
immediately after. But:</p>
|
||
<ul>
|
||
<li><code>PyNode::stop()</code> (<a href="https://gitea.tourolle.paris/dtourolle/KPN/raw/commit/4b6e498ba7e70a34cc0b57638f9e56a43b7f41ae/include/kpn/python/bindings.hpp"><code>include/kpn/python/bindings.hpp</code></a>
|
||
in the KPN++ submodule) is the <em>only</em>
|
||
code that sets <code>stop_flag_ = true</code> before joining the node's worker thread.</li>
|
||
<li>The source node's <code>run_loop()</code> has <code>while (!stop_flag_)</code> as its only exit
|
||
condition (it has no input channels, so it never sees a channel-closed signal
|
||
either).</li>
|
||
<li>Skipping <code>stop()</code> meant <code>stop_flag_</code> never became true. When <code>replay()</code> returned,
|
||
its local <code>net</code> went out of scope immediately, running <code>~PyNetwork</code> → <code>~PyNode</code> →
|
||
<code>thread_.join()</code> <strong>synchronously inside <code>replay()</code>'s own call frame</strong> — before
|
||
<code>main()</code> ever got control back to run <code>os._exit(0)</code>.</li>
|
||
</ul>
|
||
<p>Root-caused via <code>gdb -p <pid> -batch -ex "thread apply all bt"</code> on a hung process:
|
||
the main thread was stuck in <code>~PyNode</code>'s <code>jthread::join()</code>; the worker thread was in
|
||
an ordinary <code>time.sleep()</code> inside the Python source callback, waiting for a stop
|
||
signal that was never sent. The two HSA <code>kfd_wait_on_events</code> threads visible in the
|
||
same trace are normal ROCm runtime housekeeping, not evidence of a wedged GPU kernel.</p>
|
||
<p><strong>Fix:</strong> <code>replay.py</code> now calls <code>replay(..., stop=True)</code> (the removed <code>stop=False</code> +
|
||
<code>os._exit</code> workaround was actively harmful). Verified 3/3 clean runs at ~8s each
|
||
(down from a guaranteed 45s timeout), and a full DE sweep producing real, sensible
|
||
F1/precision/recall instead of flat 0.0%.</p>
|
||
<h2 id="concurrency-tuning">Concurrency tuning<a class="headerlink" href="#concurrency-tuning" title="Permanent link">¶</a></h2>
|
||
<p>With the deadlock fixed,
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/scripts/optimizer/optimize.py"><code>scripts/optimizer/optimize.py</code></a>
|
||
was extended with DE-level parallelism —
|
||
<code>differential_evolution(..., workers=ThreadPoolExecutor.map)</code> — so multiple
|
||
population candidates evaluate concurrently, each spawning its own per-film replay
|
||
subprocesses (<code>REPLAY_WORKERS</code>). Total concurrent GPU replay processes ≈
|
||
<code>DE_WORKERS × REPLAY_WORKERS</code>.</p>
|
||
<table>
|
||
<thead>
|
||
<tr>
|
||
<th>concurrent replays</th>
|
||
<th>result</th>
|
||
</tr>
|
||
</thead>
|
||
<tbody>
|
||
<tr>
|
||
<td>3 (<code>REPLAY_WORKERS=3</code>, no DE parallelism)</td>
|
||
<td>baseline, GPU underutilised</td>
|
||
</tr>
|
||
<tr>
|
||
<td>6 (<code>DE_WORKERS=2 × REPLAY_WORKERS=3</code>)</td>
|
||
<td>clean, real scores, ~1 isolated timeout per run</td>
|
||
</tr>
|
||
<tr>
|
||
<td>8 (<code>DE_WORKERS=2 × REPLAY_WORKERS=4</code>, 4-film manifest)</td>
|
||
<td>clean, real scores</td>
|
||
</tr>
|
||
<tr>
|
||
<td>9 (<code>DE_WORKERS=3 × REPLAY_WORKERS=3</code>)</td>
|
||
<td><strong>broken</strong> — every replay blew past the 45s timeout, all scores silently degraded to 0.0%</td>
|
||
</tr>
|
||
</tbody>
|
||
</table>
|
||
<p>9 concurrent replays looks like valid output (well-formed JSON, a real number) while
|
||
actually being garbage — a dangerous failure mode, not a crash. <strong>8 concurrent is the
|
||
practical ceiling</strong> on this GPU (gfx1100) for this workload. The matrix ran at
|
||
<code>REPLAY_WORKERS=4 DE_WORKERS=2</code>.</p>
|
||
<h2 id="training-films-and-validation-set">Training films and validation set<a class="headerlink" href="#training-films-and-validation-set" title="Permanent link">¶</a></h2>
|
||
<p>9 films total have dumped embeddings across all 4 models. 4 were used for
|
||
optimization, leaving 5 held out for validation:</p>
|
||
<ul>
|
||
<li><strong>Lord of War</strong> (64-cast, "clean")</li>
|
||
<li><strong>Scarface</strong> (67-cast, "ensemble/lookalike")</li>
|
||
<li><strong>Sound of Metal</strong> (14-cast, "high gallery-coverage")</li>
|
||
<li><strong>Café Society</strong> (62-cast, added this round — similar ensemble size to Scarface but
|
||
different genre/lighting; picked to add diversity, not genre-overlap, over
|
||
Downton Abbey or The Many Saints of Newark)</li>
|
||
</ul>
|
||
<p>Held out: Benny & Joon, Downton Abbey: A New Era, Lovelace, The Many Saints of
|
||
Newark, Valerian and the City of a Thousand Planets. The training-set numbers below are
|
||
training-set fit — see "Held-out validation" further down for the real
|
||
generalization test.</p>
|
||
<h2 id="search-space-and-de-settings">Search space and DE settings<a class="headerlink" href="#search-space-and-de-settings" title="Permanent link">¶</a></h2>
|
||
<p><code>popsize=10, maxiter=15</code> (3 params → ≤480 evals/combo ceiling; DE's <code>tol</code> convergence
|
||
usually stops earlier). <code>anneal_sec</code>/<code>extinction_sec</code> bounds were <strong>widened from
|
||
1–30/1–15 to 1–60/1–60 mid-run</strong> (see below) — the 4 <code>arcface_w600k_r50</code> combos
|
||
finished before the widening and still use the old, narrower bounds, so they are
|
||
<strong>not directly comparable</strong> to the other 12 on those two params. Re-running r50 with
|
||
the wider bounds was deferred (diminishing-returns judgment call, not yet done).</p>
|
||
<h2 id="results-all-16-combos-4-models-full-restricted-expand-noexp">Results — all 16 combos (4 models × {full, restricted} × {expand, noexp})<a class="headerlink" href="#results-all-16-combos-4-models-full-restricted-expand-noexp" title="Permanent link">¶</a></h2>
|
||
<p>Ranked by F1. <code>misid</code> = FPI_misid, count of true wrong-actor identifications (an
|
||
actor named who isn't in the film's cast at all) — distinct from <code>FPI</code>, which
|
||
includes in-cast timing slips.</p>
|
||
<table>
|
||
<thead>
|
||
<tr>
|
||
<th>combo</th>
|
||
<th>F1</th>
|
||
<th>P</th>
|
||
<th>R</th>
|
||
<th>TPI</th>
|
||
<th>FPI</th>
|
||
<th>misid</th>
|
||
<th>FN</th>
|
||
</tr>
|
||
</thead>
|
||
<tbody>
|
||
<tr>
|
||
<td>LVFace-B_Glint360K_restricted_exp</td>
|
||
<td><strong>78.3%</strong></td>
|
||
<td>91.0%</td>
|
||
<td>68.9%</td>
|
||
<td>42830</td>
|
||
<td>3782</td>
|
||
<td>60</td>
|
||
<td>19492</td>
|
||
</tr>
|
||
<tr>
|
||
<td>LVFace-B_Glint360K_restricted_noexp</td>
|
||
<td>76.7%</td>
|
||
<td>91.5%</td>
|
||
<td>66.2%</td>
|
||
<td>41149</td>
|
||
<td>3400</td>
|
||
<td>59</td>
|
||
<td>21173</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_w600k_mbf_restricted_exp</td>
|
||
<td>76.5%</td>
|
||
<td>90.7%</td>
|
||
<td>66.3%</td>
|
||
<td>41270</td>
|
||
<td>4431</td>
|
||
<td><strong>0</strong></td>
|
||
<td>21052</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_r18_restricted_exp</td>
|
||
<td>75.5%</td>
|
||
<td>87.6%</td>
|
||
<td>66.5%</td>
|
||
<td>41399</td>
|
||
<td>5666</td>
|
||
<td>60</td>
|
||
<td>20923</td>
|
||
</tr>
|
||
<tr>
|
||
<td><strong>LVFace-B_Glint360K_full_exp</strong></td>
|
||
<td><strong>75.3%</strong></td>
|
||
<td>89.7%</td>
|
||
<td>65.4%</td>
|
||
<td>47757</td>
|
||
<td>3407</td>
|
||
<td>232</td>
|
||
<td>26966</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_w600k_mbf_restricted_noexp</td>
|
||
<td>75.0%</td>
|
||
<td>91.1%</td>
|
||
<td>63.9%</td>
|
||
<td>39752</td>
|
||
<td>3465</td>
|
||
<td>60</td>
|
||
<td>22570</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_w600k_mbf_full_noexp</td>
|
||
<td>74.2%</td>
|
||
<td>87.4%</td>
|
||
<td>64.4%</td>
|
||
<td>12645</td>
|
||
<td>1312</td>
|
||
<td>57</td>
|
||
<td>6985</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_r18_restricted_noexp</td>
|
||
<td>73.5%</td>
|
||
<td>91.3%</td>
|
||
<td>61.7%</td>
|
||
<td>38299</td>
|
||
<td>3220</td>
|
||
<td>60</td>
|
||
<td>24023</td>
|
||
</tr>
|
||
<tr>
|
||
<td>LVFace-B_Glint360K_full_noexp</td>
|
||
<td>72.4%</td>
|
||
<td>94.2%</td>
|
||
<td>58.9%</td>
|
||
<td>27077</td>
|
||
<td>1725</td>
|
||
<td><strong>0</strong></td>
|
||
<td>19506</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_w600k_mbf_full_exp</td>
|
||
<td>72.0%</td>
|
||
<td>87.7%</td>
|
||
<td>61.4%</td>
|
||
<td>39875</td>
|
||
<td>3729</td>
|
||
<td>240</td>
|
||
<td>26338</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_w600k_r50_full_noexp †</td>
|
||
<td>71.6%</td>
|
||
<td>96.7%</td>
|
||
<td>56.9%</td>
|
||
<td>22471</td>
|
||
<td>361</td>
|
||
<td>45</td>
|
||
<td>17012</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_w600k_r50_restricted_exp †</td>
|
||
<td>71.1%</td>
|
||
<td>96.5%</td>
|
||
<td>56.4%</td>
|
||
<td>34954</td>
|
||
<td>1119</td>
|
||
<td>15</td>
|
||
<td>27368</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_w600k_r50_restricted_noexp †</td>
|
||
<td>69.2%</td>
|
||
<td>97.9%</td>
|
||
<td>53.6%</td>
|
||
<td>21146</td>
|
||
<td>327</td>
|
||
<td>15</td>
|
||
<td>18337</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_r18_full_exp</td>
|
||
<td>69.1%</td>
|
||
<td>87.6%</td>
|
||
<td>57.7%</td>
|
||
<td>37342</td>
|
||
<td>3119</td>
|
||
<td>242</td>
|
||
<td>28871</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_w600k_r50_full_exp †</td>
|
||
<td>68.5%</td>
|
||
<td>94.0%</td>
|
||
<td>54.1%</td>
|
||
<td>34982</td>
|
||
<td>903</td>
|
||
<td>150</td>
|
||
<td>31231</td>
|
||
</tr>
|
||
<tr>
|
||
<td>arcface_r18_full_noexp</td>
|
||
<td>66.6%</td>
|
||
<td>91.3%</td>
|
||
<td>53.1%</td>
|
||
<td>34314</td>
|
||
<td>2362</td>
|
||
<td>107</td>
|
||
<td>31899</td>
|
||
</tr>
|
||
</tbody>
|
||
</table>
|
||
<p>† old, narrower anneal/extinction bounds (see above) — not directly comparable to
|
||
the other 12 on those two params.</p>
|
||
<p>The same 16 results as a picture — the two headline effects are visible without
|
||
reading a single row: filled (restricted) dots stack the top of the ranking for
|
||
every model color, and yellow (LVFace) leads within both scopes:</p>
|
||
<p><img alt="All 16 bake-off combos ranked by training-set F1" src="../assets/images/rep4_matrix_f1.png" /></p>
|
||
<h2 id="calibration-curves-discriminative-power-independent-of-the-threshold">Calibration curves — discriminative power, independent of the threshold<a class="headerlink" href="#calibration-curves-discriminative-power-independent-of-the-threshold" title="Permanent link">¶</a></h2>
|
||
<p>Each model's gallery carries a fitted Platt sigmoid <code>P(match | sim) = σ(a·sim + b)</code>
|
||
(embedded directly in the gallery HDF5, see
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/src/gallery/gallery_calibration.hpp"><code>src/gallery/gallery_calibration.hpp</code></a>).
|
||
Plotting all four side by side shows discriminative power directly, independent of
|
||
whatever <code>prob_threshold</code> a particular run happened to use:</p>
|
||
<p><img alt="Calibrated P(match|similarity) for all four models" src="../assets/images/calibration_curves.png" /></p>
|
||
<p>LVFace-B has both the steepest curve (<code>a=17.7</code>, 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) —
|
||
it separates same-actor from different-actor pairs more confidently at a lower
|
||
similarity, consistent with it winning the full-gallery F1 comparison below.
|
||
Generated by
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/scripts/docs/calibration_chart.py"><code>scripts/docs/calibration_chart.py</code></a>
|
||
(requires each gallery to have been calibrated at least once — run any replay
|
||
against it first).</p>
|
||
<h2 id="two-effects-in-isolation-gallery-scope-and-pose-expansion">Two effects in isolation: gallery scope, and pose expansion<a class="headerlink" href="#two-effects-in-isolation-gallery-scope-and-pose-expansion" title="Permanent link">¶</a></h2>
|
||
<p>The matrix crosses two independent variables — averaging across all 4 models
|
||
isolates each one from model choice:</p>
|
||
<p><strong>Gallery scope (whole 2418-actor gallery vs. restricted to the film's credited
|
||
cast)</strong> — averaged over both expansion settings and all 4 models:</p>
|
||
<table>
|
||
<thead>
|
||
<tr>
|
||
<th>scope</th>
|
||
<th>F1</th>
|
||
<th>P</th>
|
||
<th>R</th>
|
||
<th>total misID (16 evals→8 each)</th>
|
||
</tr>
|
||
</thead>
|
||
<tbody>
|
||
<tr>
|
||
<td>full</td>
|
||
<td>71.2%</td>
|
||
<td>91.1%</td>
|
||
<td>59.0%</td>
|
||
<td>1073</td>
|
||
</tr>
|
||
<tr>
|
||
<td><strong>restricted</strong></td>
|
||
<td><strong>74.5%</strong></td>
|
||
<td>92.2%</td>
|
||
<td><strong>62.9%</strong></td>
|
||
<td><strong>329</strong></td>
|
||
</tr>
|
||
</tbody>
|
||
</table>
|
||
<p>Restriction wins outright on every axis — not a precision/recall trade, a clean
|
||
win: <strong>+3.3pp F1, +3.9pp recall, and less than a third the total misIDs.</strong> Fewer
|
||
candidates in the matcher's search space means fewer opportunities for a
|
||
look-alike false match, and (per the recall gain) doesn't cost real detections.
|
||
This is the single cleanest signal in the whole matrix — stronger than the model
|
||
choice itself — which is exactly why cast-restriction becoming a real runtime
|
||
feature (not just an optimizer trick) is the top item in Caveats below.</p>
|
||
<p><strong>Pose expansion (promoting a confidently-identified track's novel-pose views into
|
||
a per-film gallery annex — <a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/src/gallery/track_gallery.hpp"><code>src/gallery/track_gallery.hpp</code></a>)</strong>
|
||
is smaller and interacts with
|
||
scope rather than acting independently:</p>
|
||
<table>
|
||
<thead>
|
||
<tr>
|
||
<th>scope</th>
|
||
<th>expansion</th>
|
||
<th>F1</th>
|
||
<th>R</th>
|
||
<th>misID</th>
|
||
</tr>
|
||
</thead>
|
||
<tbody>
|
||
<tr>
|
||
<td>full</td>
|
||
<td>off</td>
|
||
<td>71.2%</td>
|
||
<td>58.3%</td>
|
||
<td>209</td>
|
||
</tr>
|
||
<tr>
|
||
<td>full</td>
|
||
<td><strong>on</strong></td>
|
||
<td>71.2%</td>
|
||
<td>59.7%</td>
|
||
<td><strong>864</strong></td>
|
||
</tr>
|
||
<tr>
|
||
<td>restricted</td>
|
||
<td>off</td>
|
||
<td>73.6%</td>
|
||
<td>61.3%</td>
|
||
<td>194</td>
|
||
</tr>
|
||
<tr>
|
||
<td>restricted</td>
|
||
<td><strong>on</strong></td>
|
||
<td><strong>75.4%</strong></td>
|
||
<td><strong>64.5%</strong></td>
|
||
<td>135</td>
|
||
</tr>
|
||
</tbody>
|
||
</table>
|
||
<p>In <strong>restricted</strong> mode, expansion is a clean win (+1.8pp F1, +3.2pp recall, misID
|
||
actually <em>drops</em>) — the annex only ever competes against the film's own ~15-actor
|
||
cast, so a "confidently identified, new pose" view is unlikely to be mistaken for
|
||
someone else. In <strong>full</strong> mode, expansion buys essentially nothing on F1 (71.2% →
|
||
71.2%, recall +1.4pp) while <strong>quadrupling misIDs</strong> (209 → 864): a novel-pose view
|
||
promoted into the annex now competes against the whole 2418-actor gallery, so a
|
||
"confident" identity is confident against the wrong universe of candidates — the
|
||
expansion mechanism is "learning" a pose correctly, but the enlarged evidence pool
|
||
makes it easier for that learned pose to look like a plausible match for a
|
||
different actor. <strong>Practical takeaway: gallery expansion should be paired with
|
||
cast restriction, not used on the full gallery</strong> — the version currently shipped
|
||
as default (<code>full_exp</code>, see TL;DR) sits in the worse of these four cells for this
|
||
specific knob, even though it's the best available combo without cast-restriction
|
||
support in the app yet (see Caveats).</p>
|
||
<h2 id="what-the-data-says">What the data says<a class="headerlink" href="#what-the-data-says" title="Permanent link">¶</a></h2>
|
||
<ul>
|
||
<li><strong>LVFace was worth its size.</strong> It wins full-gallery mode outright (75.3% vs r50's
|
||
68.5%, r18's 69.1%, mbf's 72.0%) with the highest recall of any full-mode combo —
|
||
the earlier scene-union-metric conclusion ("not worth it") doesn't survive the
|
||
better metric.</li>
|
||
<li><strong>Cast-restriction is a consistent, broad win.</strong> Every model's best combo is
|
||
<code>restricted</code>. It isn't just precision-safe: <code>arcface_w600k_mbf_restricted_exp</code> and
|
||
<code>LVFace-B_Glint360K_full_noexp</code> both hit <strong>misid=0</strong> — zero true wrong-actor
|
||
identifications. But restriction is an <strong>offline optimizer technique, not a live
|
||
app feature</strong> — it pre-filters each film's gallery to its Jellyfin-credited cast
|
||
before the matcher ever runs; there's no runtime "restrict to this film's cast"
|
||
switch in the app today. Implementing it for real is future work, tracked
|
||
separately from this defaults update.</li>
|
||
<li><strong>Gallery expansion (<code>expand_gallery</code>) is mode-dependent.</strong> It helps on
|
||
<code>restricted</code> galleries (smaller, so novel-pose promotion adds real signal) and on
|
||
LVFace's full gallery, but <strong>hurts</strong> r50 and mbf in full mode (compare
|
||
<code>arcface_w600k_r50_full_exp</code> 68.5% vs <code>full_noexp</code> 71.6%). Don't assume it's a free
|
||
win — model- and mode-dependent.</li>
|
||
<li><strong>arcface_r18 (smallest/cheapest) is last across all 4 modes</strong> — model capacity
|
||
matters here, this isn't just parameter-count padding.</li>
|
||
<li><strong><code>anneal_sec</code>/<code>extinction_sec</code> kept pinning at the search ceiling.</strong> With the
|
||
original 1–30/1–15 bounds, 3 of 4 r50 combos landed at ~93-98% of the upper bound.
|
||
Widened to 1–60/1–60 mid-run (after the r50 combos had already finished) — every
|
||
subsequent combo's best config landed at ~90%+ of the <em>new</em> ceiling too (e.g. the
|
||
LVFace winner: <code>ann=59.2, ext=59.2</code>, both ~99% of 60). The likely mechanism: a
|
||
strict <code>prob_threshold</code> "earns" a long extinction/anneal window — once false
|
||
matches are rare, a long window just bridges real presence gaps (occlusion, turned
|
||
face) instead of smearing false positives into later scenes, which is what made
|
||
short windows look better under the old, laxer thresholds. <strong>Open question, not
|
||
resolved</strong>: does this keep climbing past 60s, or does it actually plateau there?
|
||
Decided not to chase further this round (diminishing-returns judgment call) — flag
|
||
for a future sweep if it matters.</li>
|
||
</ul>
|
||
<p>The ceiling-pinning is visible in the raw search itself. Every one of the 512
|
||
DE evaluations for the winning combo, plotted over the
|
||
<code>prob_threshold</code> × <code>extinction_sec</code> plane:</p>
|
||
<p><img alt="DE search landscape: 512 evaluations over prob_threshold × extinction_sec" src="../assets/images/de_search_landscape.png" /></p>
|
||
<p>The dark band hugging the top edge <em>is</em> the finding: nearly everything scoring
|
||
well sits at <code>extinction_sec</code> ≥ 50, across a wide range of thresholds, and the
|
||
population converged into a dense cloud around the optimum (threshold ~0.70–0.80,
|
||
extinction pinned at the 60s bound). Short extinction windows (bottom half) are
|
||
uniformly pale — under a strict threshold there is simply no good configuration
|
||
down there. Generated by
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/scripts/docs/experiment_charts.py"><code>scripts/docs/experiment_charts.py</code></a>
|
||
from the DE trajectories (<code>experiments/trajectories/*.jsonl</code>, part of the
|
||
<code>experiment-data</code> artifact package).</p>
|
||
<h2 id="held-out-validation-the-number-that-actually-matters">Held-out validation — the number that actually matters<a class="headerlink" href="#held-out-validation-the-number-that-actually-matters" title="Permanent link">¶</a></h2>
|
||
<p>The 16-combo matrix above is training-set fit. This is the real test: the shipped
|
||
config (<code>LVFace-B_Glint360K_full_exp</code> — <code>prob_threshold=0.754, anneal_sec=35.5,
|
||
extinction_sec=57.4, expand_gallery=true</code>) replayed against the <strong>5 films never seen
|
||
by the optimizer</strong> (Benny & Joon, Downton Abbey: A New Era, Lovelace, The Many
|
||
Saints of Newark, Valerian and the City of a Thousand Planets), scored the same way.</p>
|
||
<table>
|
||
<thead>
|
||
<tr>
|
||
<th>film</th>
|
||
<th>F1</th>
|
||
<th>P</th>
|
||
<th>R</th>
|
||
<th>agree</th>
|
||
<th>TPI</th>
|
||
<th>FPI</th>
|
||
<th>misid</th>
|
||
<th>FN</th>
|
||
</tr>
|
||
</thead>
|
||
<tbody>
|
||
<tr>
|
||
<td>Benny & Joon</td>
|
||
<td>83.0%</td>
|
||
<td>89.1%</td>
|
||
<td>77.7%</td>
|
||
<td>72.4%</td>
|
||
<td>15125</td>
|
||
<td>1846</td>
|
||
<td>0</td>
|
||
<td>4337</td>
|
||
</tr>
|
||
<tr>
|
||
<td>Lovelace</td>
|
||
<td>77.5%</td>
|
||
<td>90.3%</td>
|
||
<td>67.9%</td>
|
||
<td>72.1%</td>
|
||
<td>14990</td>
|
||
<td>1085</td>
|
||
<td>58</td>
|
||
<td>7085</td>
|
||
</tr>
|
||
<tr>
|
||
<td>Valerian and the City of a Thousand Planets</td>
|
||
<td>74.1%</td>
|
||
<td>97.1%</td>
|
||
<td>60.0%</td>
|
||
<td>58.8%</td>
|
||
<td>18663</td>
|
||
<td>548</td>
|
||
<td>0</td>
|
||
<td>12467</td>
|
||
</tr>
|
||
<tr>
|
||
<td>Downton Abbey: A New Era</td>
|
||
<td>56.2%</td>
|
||
<td>97.8%</td>
|
||
<td>39.4%</td>
|
||
<td>40.6%</td>
|
||
<td>52027</td>
|
||
<td>1173</td>
|
||
<td>0</td>
|
||
<td>80084</td>
|
||
</tr>
|
||
<tr>
|
||
<td><strong>The Many Saints of Newark</strong></td>
|
||
<td><strong>46.3%</strong></td>
|
||
<td><strong>54.7%</strong></td>
|
||
<td>40.1%</td>
|
||
<td>37.0%</td>
|
||
<td>15922</td>
|
||
<td>4394</td>
|
||
<td><strong>974</strong></td>
|
||
<td>23791</td>
|
||
</tr>
|
||
<tr>
|
||
<td><strong>macro average (5 films)</strong></td>
|
||
<td><strong>67.4%</strong></td>
|
||
<td>85.8%</td>
|
||
<td>57.0%</td>
|
||
<td>56.2%</td>
|
||
<td>116727</td>
|
||
<td>9046</td>
|
||
<td>1032</td>
|
||
<td>127764</td>
|
||
</tr>
|
||
</tbody>
|
||
</table>
|
||
<p><img alt="Held-out per-film F1 vs. the training-set fit" src="../assets/images/holdout_f1_by_film.png" /></p>
|
||
<p><strong>67.4% held out vs. 75.3% on training</strong> — an ~8pp drop, and a much more informative
|
||
number than the training-set F1 alone: a <strong>37pp spread between best and worst film</strong>
|
||
(83.0% vs 46.3%). The config does not generalize uniformly.</p>
|
||
<p>Two films are outright failure cases, and rendering bounding boxes + names on the
|
||
extracted frames (<code>replay.py --raw-out</code> +
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/scripts/optimizer/dump_error_frames.py"><code>dump_error_frames.py</code></a><code>--raw</code>, see
|
||
Reproduce) turned what looked like a same-scene misidentification into something
|
||
more precise and more damning:</p>
|
||
<ul>
|
||
<li><strong>The Many Saints of Newark</strong> (mob-family drama, picked as a training-adjacent
|
||
genre test) has <strong>974 true misIDs</strong> — far more than any training combo saw at any
|
||
setting. The annotated frame below shows the same mechanism as Downton Abbey,
|
||
at smaller scale: <strong>"Jon Bernthal 100%", "Joey Diaz 100%", and "Billy Magnussen
|
||
100%" are all frozen boxes over empty background — a blurred wall, hanging plates —
|
||
with no face in them at all.</strong> Only one real face in frame has a box, and it
|
||
carries a <em>second</em>, colliding label ("Leslie Odom Jr." and "Michael Gandolfini"
|
||
both at high confidence on the same box) — likely two tracks whose frozen bboxes
|
||
happen to overlap.</li>
|
||
</ul>
|
||
<p><img alt="Frozen ghost boxes over background, The Many Saints of Newark" src="../assets/images/many_saints_ghost_fpi.jpg" />
|
||
<em>Frame <code>many_saints/fpi/fpi_t03543.jpg</code> from the <code>montage-frames</code> artifact
|
||
package (<code>scripts/artifacts/pull_artifacts.sh montage-frames
|
||
Many_Saints_of_Newark</code>).</em></p>
|
||
<ul>
|
||
<li><strong>Downton Abbey: A New Era</strong> (large ensemble, 36-cast) has high precision (97.8%)
|
||
but recall collapses to 39.4% (FN=80084, by far the largest of the 5). Its
|
||
starkest failure happens where there is nothing to see at all: the film's hard
|
||
cut into its closing credits, where <strong>the matcher kept reporting 15 actors —
|
||
all wrong — for nearly a minute of faceless screen.</strong></li>
|
||
</ul>
|
||
<p>Both are the same mechanism, and it can be <em>measured</em>, not just screenshotted.
|
||
Plotting the dump's own per-second <code>face_count</code> (detector output, independent
|
||
of the tracker) against the number of actors the tracker reports, through
|
||
Downton Abbey's cut to credits:</p>
|
||
<p><img alt="Detector face_count vs. tracker-reported actors through the cut to credits" src="../assets/images/downton_ghost_timeline.png" /></p>
|
||
<p>From the cut onward the detector sees <strong>zero faces</strong> — yet the tracker holds a
|
||
perfectly flat plateau of 15 reported identities for 56 seconds, each with the
|
||
<em>exact same bbox, unchanged to the pixel</em> (verified for Hugh Bonneville:
|
||
<code>(1743.2, 0.0, 171.3, 317.8)</code> at every sampled second from 7222 through 7279+).
|
||
The staircase on the right edge is the extinction window finally expiring,
|
||
actor by actor. That plateau is <code>SceneTrackerFunc</code>'s
|
||
<code>active_[actor_idx].last_bbox</code>
|
||
(<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/src/nodes/scene_tracker_node.hpp"><code>src/nodes/scene_tracker_node.hpp</code></a>)
|
||
being re-emitted
|
||
unchanged — <strong>the extinction state machine working exactly as coded</strong>, not a
|
||
bug in the logic. The film cuts from a packed group shot straight into ~40+
|
||
seconds of blank titles/credits with zero faces, and <code>extinction_sec=57.4</code> is
|
||
comfortably long enough to bridge that entire gap without expiring, so the
|
||
tracker faithfully keeps reporting "last known position" for a cast that is no
|
||
longer on screen at all.</p>
|
||
<p>This reframes the "long extinction window wins" DE-search pattern (see above): it
|
||
isn't unambiguously good. It buys recall by bridging real gaps (occlusion, turned
|
||
face) in some films, but on others — specifically, hard cuts into long faceless
|
||
footage — it manufactures a frozen-bbox ghost the tracker has no way to verify,
|
||
precisely the failure mode the <em>original</em> short-extinction-window default (<code>1.5s</code>)
|
||
was chosen to avoid. The training-set films apparently didn't have a long enough
|
||
faceless stretch after a confirmed identity to expose this; the held-out set did.</p>
|
||
<p>Frames for all three films (<code>benny_joon</code>, <code>many_saints</code>, <code>downton_abbey</code> — one strong
|
||
performer, two failure cases) are under <code>experiments/results/holdout/frames/</code>
|
||
(not committed — pull per film with <code>scripts/artifacts/pull_artifacts.sh
|
||
montage-frames <film-slug></code>), each
|
||
with a <code>manifest.json</code> listing the bucket (<code>best</code>/<code>fpi</code>/<code>fn</code>), timestamp, and
|
||
predicted vs. ground-truth actors for every dumped frame. Frames are annotated with
|
||
bounding boxes + name/confidence (green = identified, orange = unknown), matching
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/src/nodes/debug_renderer_node.hpp"><code>src/nodes/debug_renderer_node.hpp</code></a>'s
|
||
colour convention. Generated by
|
||
<code>scripts/optimizer/dump_error_frames.py --raw <replay.py --raw-out output></code> (see
|
||
Reproduce).</p>
|
||
<p><code>dump_error_frames.py --interval-sec 600</code> also supports a per-N-second sweep
|
||
instead of the fixed best/fpi/fn buckets: one best (highest Jaccard) and one worst
|
||
(lowest Jaccard) frame per 10-minute window across the whole film, e.g.
|
||
<code>experiments/results/holdout/frames/many_saints_intervals/</code> (13 windows × 2 = 26
|
||
frames for the ~2h Many Saints runtime) — a way to sample "how are we doing" evenly
|
||
across a film's runtime rather than only at its most extreme seconds.</p>
|
||
<h2 id="caveats-what-this-is-not">Caveats / what this is not<a class="headerlink" href="#caveats-what-this-is-not" title="Permanent link">¶</a></h2>
|
||
<ul>
|
||
<li><strong>r50's 4 combos used the old, narrower search bounds</strong> and aren't fully
|
||
comparable to the other 12 on <code>anneal_sec</code>/<code>extinction_sec</code>.</li>
|
||
<li><strong>The applied defaults use <code>full_exp</code>, not the higher-scoring <code>restricted_exp</code></strong>,
|
||
because cast-restriction isn't a real runtime feature yet (see above). The
|
||
78.3% F1 number is not what the shipped defaults will produce — 75.3% is.</li>
|
||
<li><strong><code>full_exp</code> is the best full-gallery combo, but not the safest.</strong> Per the
|
||
isolated-effects analysis above, <code>expand_gallery=true</code> only cleanly pays off
|
||
when paired with cast-restriction; on the full gallery it's flat on F1 while
|
||
~4x-ing misIDs (209→864, averaged across models). <code>full_noexp</code> scores lower
|
||
(72.4% vs 75.3% for LVFace) but with <strong>zero</strong> true misIDs and higher precision
|
||
(94.2% vs 89.7%). Kept <code>full_exp</code> as shipped since it's the highest-F1 option
|
||
available without cast-restriction, but this is a real F1-vs-safety trade, not
|
||
a strictly-better choice — worth revisiting if misID rate matters more than
|
||
the last few points of F1 for a given deployment.</li>
|
||
<li><strong>Switching the default model is an operational change, not just a config tweak</strong>:
|
||
any existing gallery built from r50 embeddings is incompatible with LVFace
|
||
embeddings and needs rebuilding.</li>
|
||
</ul>
|
||
<h2 id="reproduce">Reproduce<a class="headerlink" href="#reproduce" title="Permanent link">¶</a></h2>
|
||
<div class="language-bash highlight"><pre><span></span><code><span id="__span-0-1"><a id="__codelineno-0-1" name="__codelineno-0-1" href="#__codelineno-0-1"></a><span class="c1"># 4-film matrix, all 4 models × 2 modes × 2 expansion settings</span>
|
||
</span><span id="__span-0-2"><a id="__codelineno-0-2" name="__codelineno-0-2" href="#__codelineno-0-2"></a>bash<span class="w"> </span>experiments/run_rep4_subprocess.sh
|
||
</span><span id="__span-0-3"><a id="__codelineno-0-3" name="__codelineno-0-3" href="#__codelineno-0-3"></a>
|
||
</span><span id="__span-0-4"><a id="__codelineno-0-4" name="__codelineno-0-4" href="#__codelineno-0-4"></a><span class="c1"># single combo</span>
|
||
</span><span id="__span-0-5"><a id="__codelineno-0-5" name="__codelineno-0-5" href="#__codelineno-0-5"></a><span class="nv">SAE_EXPAND</span><span class="o">=</span><span class="m">1</span><span class="w"> </span><span class="nv">REPLAY_WORKERS</span><span class="o">=</span><span class="m">4</span><span class="w"> </span><span class="nv">DE_WORKERS</span><span class="o">=</span><span class="m">2</span><span class="w"> </span>python3<span class="w"> </span>scripts/optimizer/optimize.py<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-6"><a id="__codelineno-0-6" name="__codelineno-0-6" href="#__codelineno-0-6"></a><span class="w"> </span>--manifest<span class="w"> </span>experiments/manifests/rep4_LVFace-B_Glint360K_full.json<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-7"><a id="__codelineno-0-7" name="__codelineno-0-7" href="#__codelineno-0-7"></a><span class="w"> </span>--gallery<span class="w"> </span>experiments/galleries/gallery_LVFace-B_Glint360K.h5<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-8"><a id="__codelineno-0-8" name="__codelineno-0-8" href="#__codelineno-0-8"></a><span class="w"> </span>--params<span class="w"> </span>prob_threshold:0.5:0.999<span class="w"> </span>anneal_sec:1:60<span class="w"> </span>extinction_sec:1:60<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-9"><a id="__codelineno-0-9" name="__codelineno-0-9" href="#__codelineno-0-9"></a><span class="w"> </span>--popsize<span class="w"> </span><span class="m">10</span><span class="w"> </span>--maxiter<span class="w"> </span><span class="m">15</span><span class="w"> </span>--trajectory<span class="w"> </span>traj.jsonl<span class="w"> </span>--out<span class="w"> </span>best.json
|
||
</span><span id="__span-0-10"><a id="__codelineno-0-10" name="__codelineno-0-10" href="#__codelineno-0-10"></a>
|
||
</span><span id="__span-0-11"><a id="__codelineno-0-11" name="__codelineno-0-11" href="#__codelineno-0-11"></a><span class="c1"># replay the shipped config against a held-out film — --raw-out is needed to draw</span>
|
||
</span><span id="__span-0-12"><a id="__codelineno-0-12" name="__codelineno-0-12" href="#__codelineno-0-12"></a><span class="c1"># bboxes later (the merged pred.json has no per-frame bbox, only actor windows)</span>
|
||
</span><span id="__span-0-13"><a id="__codelineno-0-13" name="__codelineno-0-13" href="#__codelineno-0-13"></a>python3<span class="w"> </span>scripts/optimizer/replay.py<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-14"><a id="__codelineno-0-14" name="__codelineno-0-14" href="#__codelineno-0-14"></a><span class="w"> </span>--dump<span class="w"> </span>experiments/dumps/LVFace-B_Glint360K/dump_<slug>.h5<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-15"><a id="__codelineno-0-15" name="__codelineno-0-15" href="#__codelineno-0-15"></a><span class="w"> </span>--gallery<span class="w"> </span>experiments/galleries/gallery_LVFace-B_Glint360K.h5<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-16"><a id="__codelineno-0-16" name="__codelineno-0-16" href="#__codelineno-0-16"></a><span class="w"> </span>--out<span class="w"> </span>pred.json<span class="w"> </span>--raw-out<span class="w"> </span>raw.jsonl<span class="w"> </span>--prob-threshold<span class="w"> </span><span class="m">0</span>.754<span class="w"> </span>--anneal-sec<span class="w"> </span><span class="m">35</span>.54<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-17"><a id="__codelineno-0-17" name="__codelineno-0-17" href="#__codelineno-0-17"></a><span class="w"> </span>--extinction-sec<span class="w"> </span><span class="m">57</span>.43<span class="w"> </span>--expand-gallery
|
||
</span><span id="__span-0-18"><a id="__codelineno-0-18" name="__codelineno-0-18" href="#__codelineno-0-18"></a>
|
||
</span><span id="__span-0-19"><a id="__codelineno-0-19" name="__codelineno-0-19" href="#__codelineno-0-19"></a><span class="c1"># regenerate the report's charts (16-combo ranking, DE landscape, held-out</span>
|
||
</span><span id="__span-0-20"><a id="__codelineno-0-20" name="__codelineno-0-20" href="#__codelineno-0-20"></a><span class="c1"># per-film F1, Downton ghost timeline) from the artifacts under experiments/</span>
|
||
</span><span id="__span-0-21"><a id="__codelineno-0-21" name="__codelineno-0-21" href="#__codelineno-0-21"></a>python3<span class="w"> </span>scripts/docs/experiment_charts.py<span class="w"> </span>--out-dir<span class="w"> </span>docs/assets/images
|
||
</span><span id="__span-0-22"><a id="__codelineno-0-22" name="__codelineno-0-22" href="#__codelineno-0-22"></a>
|
||
</span><span id="__span-0-23"><a id="__codelineno-0-23" name="__codelineno-0-23" href="#__codelineno-0-23"></a><span class="c1"># dump example frames (best-agreement / FPI / FN) for visual inspection, annotated</span>
|
||
</span><span id="__span-0-24"><a id="__codelineno-0-24" name="__codelineno-0-24" href="#__codelineno-0-24"></a><span class="c1"># with bounding boxes + names (--raw is optional; omit for unannotated frames)</span>
|
||
</span><span id="__span-0-25"><a id="__codelineno-0-25" name="__codelineno-0-25" href="#__codelineno-0-25"></a>python3<span class="w"> </span>scripts/optimizer/dump_error_frames.py<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-26"><a id="__codelineno-0-26" name="__codelineno-0-26" href="#__codelineno-0-26"></a><span class="w"> </span>--pred<span class="w"> </span>pred.json<span class="w"> </span>--raw<span class="w"> </span>raw.jsonl<span class="w"> </span>--xray<span class="w"> </span>experiments/xray/.../<xray_dir><span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-27"><a id="__codelineno-0-27" name="__codelineno-0-27" href="#__codelineno-0-27"></a><span class="w"> </span>--movie<span class="w"> </span><span class="s2">"<path to source video>"</span><span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-28"><a id="__codelineno-0-28" name="__codelineno-0-28" href="#__codelineno-0-28"></a><span class="w"> </span>--gallery<span class="w"> </span>experiments/galleries/gallery_LVFace-B_Glint360K.h5<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-29"><a id="__codelineno-0-29" name="__codelineno-0-29" href="#__codelineno-0-29"></a><span class="w"> </span>--out-dir<span class="w"> </span>experiments/results/holdout/frames/<name><span class="w"> </span>--n-per-bucket<span class="w"> </span><span class="m">4</span>
|
||
</span><span id="__span-0-30"><a id="__codelineno-0-30" name="__codelineno-0-30" href="#__codelineno-0-30"></a>
|
||
</span><span id="__span-0-31"><a id="__codelineno-0-31" name="__codelineno-0-31" href="#__codelineno-0-31"></a><span class="c1"># or: one best + one worst frame per 10-minute window across the whole film</span>
|
||
</span><span id="__span-0-32"><a id="__codelineno-0-32" name="__codelineno-0-32" href="#__codelineno-0-32"></a>python3<span class="w"> </span>scripts/optimizer/dump_error_frames.py<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-33"><a id="__codelineno-0-33" name="__codelineno-0-33" href="#__codelineno-0-33"></a><span class="w"> </span>--pred<span class="w"> </span>pred.json<span class="w"> </span>--raw<span class="w"> </span>raw.jsonl<span class="w"> </span>--xray<span class="w"> </span>experiments/xray/.../<xray_dir><span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-34"><a id="__codelineno-0-34" name="__codelineno-0-34" href="#__codelineno-0-34"></a><span class="w"> </span>--movie<span class="w"> </span><span class="s2">"<path to source video>"</span><span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-35"><a id="__codelineno-0-35" name="__codelineno-0-35" href="#__codelineno-0-35"></a><span class="w"> </span>--gallery<span class="w"> </span>experiments/galleries/gallery_LVFace-B_Glint360K.h5<span class="w"> </span><span class="se">\</span>
|
||
</span><span id="__span-0-36"><a id="__codelineno-0-36" name="__codelineno-0-36" href="#__codelineno-0-36"></a><span class="w"> </span>--out-dir<span class="w"> </span>experiments/results/holdout/frames/<name>_intervals<span class="w"> </span>--interval-sec<span class="w"> </span><span class="m">600</span>
|
||
</span></code></pre></div>
|
||
<p>See also: <a href="../optimizer-experiments/">the prior optimizer round</a> (superseded
|
||
metric) and the session log
|
||
<a href="https://gitea.tourolle.paris/dtourolle/scene-actor-extraction/raw/commit/4b5557974bef8783bacc375c0869e8f589d1b0a3/experiments/SESSION_STATE.md"><code>experiments/SESSION_STATE.md</code></a>.</p>
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
|
||
</article>
|
||
</div>
|
||
|
||
|
||
<script>var target=document.getElementById(location.hash.slice(1));target&&target.name&&(target.checked=target.name.startsWith("__tabbed_"))</script>
|
||
</div>
|
||
|
||
<button type="button" class="md-top md-icon" data-md-component="top" hidden>
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M13 20h-2V8l-5.5 5.5-1.42-1.42L12 4.16l7.92 7.92-1.42 1.42L13 8z"/></svg>
|
||
Back to top
|
||
</button>
|
||
|
||
</main>
|
||
|
||
<footer class="md-footer">
|
||
|
||
|
||
|
||
<nav class="md-footer__inner md-grid" aria-label="Footer" >
|
||
|
||
|
||
<a href="../lvface-deep-dive/" class="md-footer__link md-footer__link--prev" aria-label="Previous: LVFace Deep Dive">
|
||
<div class="md-footer__button md-icon">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M20 11v2H8l5.5 5.5-1.42 1.42L4.16 12l7.92-7.92L13.5 5.5 8 11z"/></svg>
|
||
</div>
|
||
<div class="md-footer__title">
|
||
<span class="md-footer__direction">
|
||
Previous
|
||
</span>
|
||
<div class="md-ellipsis">
|
||
LVFace Deep Dive
|
||
</div>
|
||
</div>
|
||
</a>
|
||
|
||
|
||
|
||
<a href="../optimizer-experiments/" class="md-footer__link md-footer__link--next" aria-label="Next: Optimizer Experiments (prior round)">
|
||
<div class="md-footer__title">
|
||
<span class="md-footer__direction">
|
||
Next
|
||
</span>
|
||
<div class="md-ellipsis">
|
||
Optimizer Experiments (prior round)
|
||
</div>
|
||
</div>
|
||
<div class="md-footer__button md-icon">
|
||
|
||
<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M4 11v2h12l-5.5 5.5 1.42 1.42L19.84 12l-7.92-7.92L10.5 5.5 16 11z"/></svg>
|
||
</div>
|
||
</a>
|
||
|
||
</nav>
|
||
|
||
|
||
<div class="md-footer-meta md-typeset">
|
||
<div class="md-footer-meta__inner md-grid">
|
||
<div class="md-copyright">
|
||
|
||
|
||
Made with
|
||
<a href="https://squidfunk.github.io/mkdocs-material/" target="_blank" rel="noopener">
|
||
Material for MkDocs
|
||
</a>
|
||
|
||
</div>
|
||
|
||
</div>
|
||
</div>
|
||
</footer>
|
||
|
||
</div>
|
||
<div class="md-dialog" data-md-component="dialog">
|
||
<div class="md-dialog__inner md-typeset"></div>
|
||
</div>
|
||
|
||
|
||
|
||
|
||
|
||
<script id="__config" type="application/json">{"annotate": null, "base": "..", "features": ["navigation.tabs", "navigation.sections", "navigation.top", "navigation.footer", "content.code.copy", "content.code.annotate"], "search": "../assets/javascripts/workers/search.2c215733.min.js", "tags": null, "translations": {"clipboard.copied": "Copied to clipboard", "clipboard.copy": "Copy to clipboard", "search.result.more.one": "1 more on this page", "search.result.more.other": "# more on this page", "search.result.none": "No matching documents", "search.result.one": "1 matching document", "search.result.other": "# matching documents", "search.result.placeholder": "Type to start searching", "search.result.term.missing": "Missing", "select.version": "Select version"}, "version": null}</script>
|
||
|
||
|
||
<script src="../assets/javascripts/bundle.d7400e89.min.js"></script>
|
||
|
||
|
||
</body>
|
||
</html> |