1 2<!doctype html> 3<html lang="en" class="no-js"> 4 <head> 5 6 <meta charset="utf-8"> 7 <meta name="viewport" content="width=device-width,initial-scale=1"> 8 9 <meta name="description" content="Python Facial Expression Analysis Toolbox â detect and analyze facial action units, emotions, landmarks, identity, and head pose from images and videos. 10"> 11 12 13 14 <link rel="canonical" href="https://py-feat.org/pages/faqs/"> 15 16 17 <link rel="prev" href="../au_reference/"> 18 19 20 <link rel="next" href="../../basic_tutorials/Detecting_Images/"> 21 22 23 24 25 26 <link rel="icon" href="../../images/logo/favicon.ico"> 27 <meta name="generator" content="mkdocs-1.6.1, mkdocs-material-9.7.7"> 28 29 30 31 <title>Faqs - Py-Feat</title> 32 33 34 35 <link rel="stylesheet" href="../../assets/stylesheets/main.ec1eaa64.min.css"> 36 37 38 <link rel="stylesheet" href="../../assets/stylesheets/palette.ab4e12ef.min.css"> 39 40 41 42 43 44 45 46 47 48 49 50 51 <link rel="preconnect" href="https://fonts.gstatic.com" crossorigin> 52 <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"> 53 <style>:root{--md-text-font:"Roboto";--md-code-font:"Roboto Mono"}</style> 54 55 56 57 <link rel="stylesheet" href="../../assets/_mkdocstrings.css"> 58 59 <link rel="stylesheet" href="../../stylesheets/extra.css?v=0.1.26"> 60 61 <link rel="stylesheet" href="../../stylesheets/logo_sidebar.css?v=0.1.26"> 62 63
63<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>
63 64 65 66 67 68 69 </head> 70 71 72 73 74 75 76 77 78 79 <body dir="ltr" data-md-color-scheme="default" data-md-color-primary="indigo" data-md-color-accent="indigo"> 80 81 82 <input class="md-toggle" data-md-toggle="drawer" type="checkbox" id="__drawer" autocomplete="off"> 83 <input class="md-toggle" data-md-toggle="search" type="checkbox" id="__search" autocomplete="off"> 84 <label class="md-overlay" for="__drawer"></label> 85 <div data-md-component="skip"> 86 87 88 <a href="#faqs-tips-known-issues" class="md-skip"> 89 Skip to content 90 </a> 91 92 </div> 93 <div data-md-component="announce"> 94 95 </div> 96 97 98 99 100 101 102<header class="md-header md-header--shadow" data-md-component="header"> 103 <nav class="md-header__inner md-grid" aria-label="Header"> 104 <a href="../.." title="Py-Feat" class="md-header__button md-logo" aria-label="Py-Feat" data-md-component="logo"> 105 106 <img src="../../images/logo/pyfeat_logo_green.svg" alt="logo"> 107 108 </a> 109 <label class="md-header__button md-icon" for="__drawer"> 110 111 <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M3 6h18v2H3zm0 5h18v2H3zm0 5h18v2H3z"/></svg> 112 </label> 113 <div class="md-header__title" data-md-component="header-title"> 114 <div class="md-header__ellipsis"> 115 <div class="md-header__topic"> 116 <span class="md-ellipsis"> 117 Py-Feat 118 </span> 119 </div> 120 <div class="md-header__topic" data-md-component="header-topic"> 121 <span class="md-ellipsis"> 122 123 Faqs 124 125 </span> 126 </div> 127 </div> 128 </div> 129 130 131 <form class="md-header__option" data-md-component="palette"> 132 133 134 135 136 <input class="md-option" data-md-color-media="" data-md-color-scheme="default" data-md-color-primary="indigo" data-md-color-accent="indigo" aria-label="Switch to dark mode" type="radio" name="__palette" id="__palette_0"> 137 138 <label class="md-header__button md-icon" title="Switch to dark mode" for="__palette_1" hidden> 139 <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> 140 </label> 141 142 143 144 145 146 <input class="md-option" data-md-color-media="" data-md-color-scheme="slate" data-md-color-primary="indigo" data-md-color-accent="indigo" aria-label="Switch to light mode" type="radio" name="__palette" id="__palette_1"> 147 148 <label class="md-header__button md-icon" title="Switch to light mode" for="__palette_0" hidden> 149 <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> 150 </label> 151 152 153</form> 154 155 156 157
157<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>
157 158 159 160 161 162 163 <label class="md-header__button md-icon" for="__search"> 164 165 <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> 166 </label> 167 <div class="md-search" data-md-component="search" role="dialog"> 168 <label class="md-search__overlay" for="__search"></label> 169 <div class="md-search__inner" role="search"> 170 <form class="md-search__form" name="search"> 171 <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> 172 <label class="md-search__icon md-icon" for="__search"> 173 174 <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> 175 176 <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> 177 </label> 178 <nav class="md-search__options" aria-label="Search"> 179 180 <a href="javascript:void(0)" class="md-search__icon md-icon" title="Share" aria-label="Share" data-clipboard data-clipboard-text="" data-md-component="search-share" tabindex="-1"> 181 182 <svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24"><path d="M18 16.08c-.76 0-1.44.3-1.96.77L8.91 12.7c.05-.23.09-.46.09-.7s-.04-.47-.09-.7l7.05-4.11c.54.5 1.25.81 2.04.81a3 3 0 0 0 3-3 3 3 0 0 0-3-3 3 3 0 0 0-3 3c0 .24.04.47.09.7L8.04 9.81C7.5 9.31 6.79 9 6 9a3 3 0 0 0-3 3 3 3 0 0 0 3 3c.79 0 1.5-.31 2.04-.81l7.12 4.15c-.05.21-.08.43-.08.66 0 1.61 1.31 2.91 2.92 2.91s2.92-1.3 2.92-2.91A2.92 2.92 0 0 0 18 16.08"/></svg> 183 </a> 184 185 <button type="reset" class="md-search__icon md-icon" title="Clear" aria-label="Clear" tabindex="-1"> 186 187 <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> 188 </button> 189 </nav> 190 191 <div class="md-search__suggest" data-md-component="search-suggest"></div> 192 193 </form> 194 <div class="md-search__output"> 195 <div class="md-search__scrollwrap" tabindex="0" data-md-scrollfix> 196 <div class="md-search-result" data-md-component="search-result"> 197 <div class="md-search-result__meta"> 198 Initializing search 199 </div> 200 <ol class="md-search-result__list" role="presentation"></ol> 201 </div> 202 </div> 203 </div> 204 </div> 205</div> 206 207 208 209 <div class="md-header__source"> 210 <a href="https://github.com/cosanlab/py-feat" title="Go to repository" class="md-source" data-md-component="source"> 211 <div class="md-source__icon md-icon"> 212 213 <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> 214 </div> 215 <div class="md-source__repository"> 216 GitHub 217 </div> 218</a> 219 </div> 220 221 </nav> 222 223</header> 224 225 <div class="md-container" data-md-component="container"> 226 227 228 229 230 231 232 <main class="md-main" data-md-component="main"> 233 <div class="md-main__inner md-grid"> 234 235 236 237 <div class="md-sidebar md-sidebar--primary" data-md-component="sidebar" data-md-type="navigation" > 238 <div class="md-sidebar__scrollwrap"> 239 <div class="md-sidebar__inner"> 240 241 242 243 244<nav class="md-nav md-nav--primary" aria-label="Navigation" data-md-level="0"> 245 <label class="md-nav__title" for="__drawer"> 246 <a href="../.." title="Py-Feat" class="md-nav__button md-logo" aria-label="Py-Feat" data-md-component="logo"> 247 248 <img src="../../images/logo/pyfeat_logo_green.svg" alt="logo"> 249 250 </a> 251 Py-Feat 252 </label> 253 254 <div class="md-nav__source"> 255 <a href="https://github.com/cosanlab/py-feat" title="Go to repository" class="md-source" data-md-component="source"> 256 <div class="md-source__icon md-icon"> 257 258 <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> 259 </div> 260 <div class="md-source__repository"> 261 GitHub 262 </div> 263</a> 264 </div> 265 266 <ul class="md-nav__list" data-md-scrollfix> 267 268 269 270 271 272 273 274 <li class="md-nav__item"> 275 <a href="../.." class="md-nav__link"> 276 277 278 279 <span class="md-ellipsis"> 280 281 282 Home 283 284 285 286 </span> 287 288 289 290 </a> 291 </li> 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 <li class="md-nav__item md-nav__item--active md-nav__item--section md-nav__item--nested"> 324 325 326 327 <input class="md-nav__toggle md-toggle " type="checkbox" id="__nav_2" checked> 328 329 330 <label class="md-nav__link" for="__nav_2" id="__nav_2_label" tabindex=""> 331 332 333 334 <span class="md-ellipsis"> 335 336 337 Getting Started 338 339 340 341 </span> 342 343 344
345 <span class="md-nav__icon md-icon"></span> 346 </label> 347 348 <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_2_label" aria-expanded="true"> 349 <label class="md-nav__title" for="__nav_2"> 350 <span class="md-nav__icon md-icon"></span> 351 352 353 Getting Started 354 355 356 </label> 357 <ul class="md-nav__list" data-md-scrollfix> 358 359 360 361 362 363 364 365 <li class="md-nav__item"> 366 <a href="../installation/" class="md-nav__link"> 367 368 369 370 <span class="md-ellipsis"> 371 372 373 Installation 374 375 376 377 </span> 378 379 380 381 </a> 382 </li> 383 384 385 386 387 388 389 390 391 392 393 <li class="md-nav__item"> 394 <a href="../models/" class="md-nav__link"> 395 396 397 398 <span class="md-ellipsis"> 399 400 401 Models 402 403 404 405 </span> 406 407 408 409 </a> 410 </li> 411 412 413 414 415 416 417 418 419 420 421 <li class="md-nav__item"> 422 <a href="../au_reference/" class="md-nav__link"> 423 424 425 426 <span class="md-ellipsis"> 427 428 429 AU Reference 430 431 432 433 </span> 434 435 436 437 </a> 438 </li> 439 440 441 442 443 444 445 446 447 448 449 450 451 <li class="md-nav__item md-nav__item--active"> 452 453 <input class="md-nav__toggle md-toggle" type="checkbox" id="__toc"> 454 455 456 457 458 459 <label class="md-nav__link md-nav__link--active" for="__toc"> 460 461 462 463 <span class="md-ellipsis"> 464 465 466 Faqs 467 468 469 470 </span> 471 472 473 474 <span class="md-nav__icon md-icon"></span> 475 </label> 476 477 <a href="./" class="md-nav__link md-nav__link--active"> 478 479 480 481 <span class="md-ellipsis"> 482 483 484 Faqs 485 486 487 488 </span> 489 490 491 492 </a> 493 494 495 496<nav class="md-nav md-nav--secondary" aria-label="Table of contents"> 497 498 499 500 501 502 503 <label class="md-nav__title" for="__toc"> 504 <span class="md-nav__icon md-icon"></span> 505 Table of contents 506 </label> 507 <ul class="md-nav__list" data-md-component="toc" data-md-scrollfix> 508 509 <li class="md-nav__item"> 510 <a href="#common-questions" class="md-nav__link"> 511 <span class="md-ellipsis"> 512 513 Common questions 514 515 </span> 516 </a> 517 518 <nav class="md-nav" aria-label="Common questions"> 519 <ul class="md-nav__list"> 520 521 <li class="md-nav__item"> 522 <a href="#py-feat-is-detecting-multiple-faces-in-an-image-with-a-single-face" class="md-nav__link"> 523 <span class="md-ellipsis"> 524 525 Py-feat is detecting multiple faces in an image with a single face 526 527 </span> 528 </a> 529 530</li> 531 532 <li class="md-nav__item"> 533 <a href="#in-what-order-are-detected-faces-returned" class="md-nav__link"> 534 <span class="md-ellipsis"> 535 536 In what order are detected faces returned? 537 538 </span> 539 </a> 540 541</li> 542 543 <li class="md-nav__item"> 544 <a href="#py-feat-is-treating-the-same-person-as-multiple-identities-or-treating-different-people-as-the-same-identity" class="md-nav__link"> 545 <span class="md-ellipsis"> 546 547 Py-feat is treating the same person as multiple identities or treating different people as the same identity 548 549 </span> 550 </a> 551 552</li> 553 554 <li class="md-nav__item"> 555 <a href="#how-can-i-speed-things-up-and-control-memory-usage" class="md-nav__link"> 556 <span class="md-ellipsis"> 557 558 How can I speed things up and control memory usage? 559 560 </span> 561 </a> 562 563</li> 564 565 <li class="md-nav__item"> 566 <a href="#how-fast-is-video-processing" class="md-nav__link"> 567 <span class="md-ellipsis"> 568 569 How fast is video processing? 570 571 </span> 572 </a> 573 574</li> 575 576 </ul> 577 </nav> 578 579</li> 580 581 <li class="md-nav__item"> 582 <a href="#known-issues" class="md-nav__link"> 583 <span class="md-ellipsis"> 584 585 Known issues 586 587 </span> 588 </a> 589 590</li> 591 592 <li class="md-nav__item"> 593 <a href="#community-getting-help" class="md-nav__link"> 594 <span class="md-ellipsis"> 595 596 Community & getting help 597 598 </span> 599 </a> 600 601</li> 602 603 </ul> 604 605</nav> 606 607 </li> 608 609 610 611 612 </ul> 613 </nav> 614 615 </li> 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 <li class="md-nav__item md-nav__item--section md-nav__item--nested"> 648 649 650 651 652 653 <input class="md-nav__toggle md-toggle md-toggle--indeterminate" type="checkbox" id="__nav_3" > 654 655 656 <label class="md-nav__link" for="__nav_3" id="__nav_3_label" tabindex=""> 657 658 659 660 <span class="md-ellipsis"> 661 662 663 Tutorials 664 665 666 667 </span> 668 669 670
671 <span class="md-nav__icon md-icon"></span> 672 </label> 673 674 <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_3_label" aria-expanded="false"> 675 <label class="md-nav__title" for="__nav_3"> 676 <span class="md-nav__icon md-icon"></span> 677 678 679 Tutorials 680 681 682 </label> 683 <ul class="md-nav__list" data-md-scrollfix> 684 685 686 687 688 689 690 691 <li class="md-nav__item"> 692 <a href="../../basic_tutorials/Detecting_Images/" class="md-nav__link"> 693 694 695 696 <span class="md-ellipsis"> 697 698 699 Detecting Images 700 701 702 703 </span> 704 705 706 707 </a> 708 </li> 709 710 711 712 713 714 715 716 717 718 719 <li class="md-nav__item"> 720 <a href="../../basic_tutorials/Detecting_Videos/" class="md-nav__link"> 721 722 723 724 <span class="md-ellipsis"> 725 726 727 Detecting Videos 728 729 730 731 </span> 732 733 734 735 </a> 736 </li> 737 738 739 740 741 742 743 744 745 746 747 <li class="md-nav__item"> 748 <a href="../../basic_tutorials/Plotting/" class="md-nav__link"> 749 750 751 752 <span class="md-ellipsis"> 753 754 755 Plotting 756 757 758 759 </span> 760 761 762 763 </a> 764 </li> 765 766 767 768 769 770 771 772 773 774 775 <li class="md-nav__item"> 776 <a href="../../basic_tutorials/Analysis/" class="md-nav__link"> 777 778 779 780 <span class="md-ellipsis"> 781 782 783 Analysis 784 785 786 787 </span> 788 789 790 791 </a> 792 </li> 793 794 795 796 797 798 799 800 801 802 803 <li class="md-nav__item"> 804 <a href="../../basic_tutorials/Performance/" class="md-nav__link"> 805 806 807 808 <span class="md-ellipsis"> 809 810 811 Performance 812 813 814 815 </span> 816 817 818 819 </a> 820 </li> 821 822 823 824 825 </ul> 826 </nav> 827 828 </li> 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 <li class="md-nav__item md-nav__item--section md-nav__item--nested"> 855 856 857 858 859 860 <input class="md-nav__toggle md-toggle md-toggle--indeterminate" type="checkbox" id="__nav_4" > 861 862 863 <label class="md-nav__link" for="__nav_4" id="__nav_4_label" tabindex=""> 864 865 866 867 <span class="md-ellipsis"> 868 869 870 Py-feat Live 871 872 873 874 </span> 875 876 877 878 <span class="md-nav__icon md-icon"></span> 879 </label> 880 881 <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_4_label" aria-expanded="false"> 882 <label class="md-nav__title" for="__nav_4"> 883 <span class="md-nav__icon md-icon"></span> 884 885 886 Py-feat Live 887 888 889 </label> 890 <ul class="md-nav__list" data-md-scrollfix> 891 892 893 894 895 896 897 898 <li class="md-nav__item"> 899 <a href="../pyfeat_live/" class="md-nav__link"> 900 901 902 903 <span class="md-ellipsis"> 904 905 906 Pyfeat live 907 908 909 910 </span> 911 912 913 914 </a> 915 </li> 916 917 918 919 920 921 922 923 924 925 926 <li class="md-nav__item"> 927 <a href="../pyfeatlive_changelog/" class="md-nav__link"> 928 929 930 931 <span class="md-ellipsis"> 932 933 934 Pyfeatlive changelog 935 936 937 938 </span> 939 940 941 942 </a> 943 </li> 944 945 946 947 948 </ul> 949 </nav> 950 951 </li> 952 953 954 955 956 957 958 959 960 961 962 963 964 965 966 967 968 969 970 971 972 973 974 975 976 977 <li class="md-nav__item md-nav__item--section md-nav__item--nested"> 978 979 980 981 982 983 <input class="md-nav__toggle md-toggle md-toggle--indeterminate" type="checkbox" id="__nav_5" > 984 985 986 <label class="md-nav__link" for="__nav_5" id="__nav_5_label" tabindex=""> 987 988 989 990 <span class="md-ellipsis"> 991 992 993 Benchmarks 994 995 996 997 </span> 998 999 1000
1001 <span class="md-nav__icon md-icon"></span> 1002 </label> 1003 1004 <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_5_label" aria-expanded="false"> 1005 <label class="md-nav__title" for="__nav_5"> 1006 <span class="md-nav__icon md-icon"></span> 1007 1008 1009 Benchmarks 1010 1011 1012 </label> 1013 <ul class="md-nav__list" data-md-scrollfix> 1014 1015 1016 1017 1018 1019 1020 1021 <li class="md-nav__item"> 1022 <a href="../../benchmarks/Speed/" class="md-nav__link"> 1023 1024 1025 1026 <span class="md-ellipsis"> 1027 1028 1029 Speed 1030 1031 1032 1033 </span> 1034 1035 1036 1037 </a> 1038 </li> 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 <li class="md-nav__item"> 1050 <a href="../../benchmarks/accuracy/" class="md-nav__link"> 1051 1052 1053 1054 <span class="md-ellipsis"> 1055 1056 1057 Accuracy 1058 1059 1060 1061 </span> 1062 1063 1064 1065 </a> 1066 </li> 1067 1068 1069 1070 1071 </ul> 1072 </nav> 1073 1074 </li> 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 <li class="md-nav__item md-nav__item--section md-nav__item--nested"> 1105 1106 1107 1108 1109 1110 <input class="md-nav__toggle md-toggle md-toggle--indeterminate" type="checkbox" id="__nav_6" > 1111 1112 1113 <label class="md-nav__link" for="__nav_6" id="__nav_6_label" tabindex=""> 1114 1115 1116 1117 <span class="md-ellipsis"> 1118 1119 1120 Contributing 1121 1122 1123 1124 </span> 1125 1126 1127 1128 <span class="md-nav__icon md-icon"></span> 1129 </label> 1130 1131 <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_6_label" aria-expanded="false"> 1132 <label class="md-nav__title" for="__nav_6"> 1133 <span class="md-nav__icon md-icon"></span> 1134 1135 1136 Contributing 1137 1138 1139 </label> 1140 <ul class="md-nav__list" data-md-scrollfix> 1141 1142 1143 1144 1145 1146 1147 1148 <li class="md-nav__item"> 1149 <a href="../contribute/" class="md-nav__link"> 1150 1151 1152 1153 <span class="md-ellipsis"> 1154 1155 1156 Contribute 1157 1158 1159 1160 </span> 1161 1162 1163 1164 </a> 1165 </li> 1166 1167 1168 1169 1170 1171 1172 1173 1174 1175 1176 <li class="md-nav__item"> 1177 <a href="../modelContribution/" class="md-nav__link"> 1178 1179 1180 1181 <span class="md-ellipsis"> 1182 1183 1184 modelContribution 1185 1186 1187 1188 </span> 1189 1190 1191 1192 </a> 1193 </li> 1194 1195 1196 1197 1198 1199 1200 1201 1202 1203 1204 <li class="md-nav__item"> 1205 <a href="../changelog/" class="md-nav__link"> 1206 1207 1208 1209 <span class="md-ellipsis"> 1210 1211 1212 Changelog 1213 1214 1215 1216 </span> 1217 1218 1219 1220 </a> 1221 </li> 1222 1223 1224 1225 1226 1227 1228 1229 1230 1231 1232 <li class="md-nav__item"> 1233 <a href="https://github.com/cosanlab/py-feat" class="md-nav__link"> 1234 1235 1236 1237 <span class="md-ellipsis"> 1238 1239 1240 GitHub Repository 1241 1242 1243 1244 </span> 1245 1246 1247 1248 </a> 1249 </li> 1250 1251 1252 1253 1254 </ul> 1255 </nav> 1256 1257 </li> 1258 1259 1260 1261 1262 1263 1264 1265 1266 1267 1268 1269 1270 1271 1272 1273 1274 1275 1276 1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290 1291 1292 1293 <li class="md-nav__item md-nav__item--section md-nav__item--nested"> 1294 1295 1296 1297 1298 1299 <input class="md-nav__toggle md-toggle md-toggle--indeterminate" type="checkbox" id="__nav_7" > 1300 1301 1302 <label class="md-nav__link" for="__nav_7" id="__nav_7_label" tabindex=""> 1303 1304 1305 1306 <span class="md-ellipsis"> 1307 1308 1309 API Reference 1310 1311 1312 1313 </span> 1314 1315 1316 1317 <span class="md-nav__icon md-icon"></span> 1318 </label> 1319 1320 <nav class="md-nav" data-md-level="1" aria-labelledby="__nav_7_label" aria-expanded="false"> 1321 <label class="md-nav__title" for="__nav_7">
1322 <span class="md-nav__icon md-icon"></span> 1323 1324 1325 API Reference 1326 1327 1328 </label> 1329 <ul class="md-nav__list" data-md-scrollfix> 1330 1331 1332 1333 1334 1335 1336 1337 <li class="md-nav__item"> 1338 <a href="../../api/feat/detector/" class="md-nav__link"> 1339 1340 1341 1342 <span class="md-ellipsis"> 1343 1344 1345 feat.detector 1346 1347 1348 1349 </span> 1350 1351 1352 1353 </a> 1354 </li> 1355 1356 1357 1358 1359 1360 1361 1362 1363 1364 1365 <li class="md-nav__item"> 1366 <a href="../../api/feat/detector_v2/" class="md-nav__link"> 1367 1368 1369 1370 <span class="md-ellipsis"> 1371 1372 1373 feat.detector_v2 1374 1375 1376 1377 </span> 1378 1379 1380 1381 </a> 1382 </li> 1383 1384 1385 1386 1387 1388 1389 1390 1391 1392 1393 <li class="md-nav__item"> 1394 <a href="../../api/feat/data/" class="md-nav__link"> 1395 1396 1397 1398 <span class="md-ellipsis"> 1399 1400 1401 feat.data 1402 1403 1404 1405 </span> 1406 1407 1408 1409 </a> 1410 </li> 1411 1412 1413 1414 1415 1416 1417 1418 1419 1420 1421 <li class="md-nav__item"> 1422 <a href="../../api/feat/plotting/" class="md-nav__link"> 1423 1424 1425 1426 <span class="md-ellipsis"> 1427 1428 1429 feat.plotting 1430 1431 1432 1433 </span> 1434 1435 1436 1437 </a> 1438 </li> 1439 1440 1441 1442 1443 1444 1445 1446 1447 1448 1449 <li class="md-nav__item"> 1450 <a href="../../api/feat/pretrained/" class="md-nav__link"> 1451 1452 1453 1454 <span class="md-ellipsis"> 1455 1456 1457 feat.pretrained 1458 1459 1460 1461 </span> 1462 1463 1464 1465 </a> 1466 </li> 1467 1468 1469 1470 1471 1472 1473 1474 1475 1476 1477 1478 1479 1480 1481 1482 1483 1484 1485 1486 1487 1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 <li class="md-nav__item md-nav__item--nested"> 1515 1516 1517 1518 1519 1520 <input class="md-nav__toggle md-toggle md-toggle--indeterminate" type="checkbox" id="__nav_7_6" > 1521 1522 1523 <div class="md-nav__link md-nav__container"> 1524 <a href="../../api/feat/utils/" class="md-nav__link "> 1525 1526 1527 1528 <span class="md-ellipsis"> 1529 1530 1531 feat.utils 1532 1533 1534 1535 </span> 1536 1537 1538 1539 </a> 1540 1541 1542 <label class="md-nav__link " for="__nav_7_6" id="__nav_7_6_label" tabindex="0"> 1543 <span class="md-nav__icon md-icon"></span> 1544 </label> 1545 1546 </div> 1547 1548 <nav class="md-nav" data-md-level="2" aria-labelledby="__nav_7_6_label" aria-expanded="false"> 1549 <label class="md-nav__title" for="__nav_7_6"> 1550 <span class="md-nav__icon md-icon"></span> 1551 1552 1553 feat.utils 1554 1555 1556 </label> 1557 <ul class="md-nav__list" data-md-scrollfix> 1558 1559 1560 1561 1562 1563 1564 1565 1566 1567 <li class="md-nav__item"> 1568 <a href="../../api/feat/utils/blendshape_to_au/" class="md-nav__link"> 1569 1570 1571 1572 <span class="md-ellipsis"> 1573 1574 1575 blendshape_to_au 1576 1577 1578 1579 </span> 1580 1581 1582 1583 </a> 1584 </li> 1585 1586 1587 1588 1589 1590 1591 1592 1593 1594 1595 <li class="md-nav__item"> 1596 <a href="../../api/feat/utils/face_gaze/" class="md-nav__link"> 1597 1598 1599 1600 <span class="md-ellipsis"> 1601 1602 1603 face_gaze 1604 1605 1606 1607 </span> 1608 1609 1610 1611 </a> 1612 </li> 1613 1614 1615 1616 1617 1618 1619 1620 1621 1622 1623 <li class="md-nav__item"> 1624 <a href="../../api/feat/utils/face_mask/" class="md-nav__link"> 1625 1626 1627 1628 <span class="md-ellipsis"> 1629 1630 1631 face_mask 1632 1633 1634 1635 </span> 1636 1637 1638 1639 </a> 1640 </li> 1641 1642 1643 1644 1645 1646 1647 1648 1649 1650 1651 <li class="md-nav__item"> 1652 <a href="../../api/feat/utils/face_pose/" class="md-nav__link"> 1653 1654 1655 1656 <span class="md-ellipsis"> 1657 1658 1659 face_pose 1660 1661 1662 1663 </span> 1664 1665 1666 1667 </a> 1668 </li> 1669 1670 1671 1672 1673 1674 1675 1676 1677 1678 1679 <li class="md-nav__item"> 1680 <a href="../../api/feat/utils/face_pose_mlp/" class="md-nav__link"> 1681 1682 1683 1684 <span class="md-ellipsis"> 1685 1686 1687 face_pose_mlp 1688 1689 1690 1691 </span> 1692 1693 1694 1695 </a> 1696 </li> 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 <li class="md-nav__item"> 1708 <a href="../../api/feat/utils/geometry/" class="md-nav__link"> 1709 1710 1711 1712 <span class="md-ellipsis"> 1713 1714 1715 geometry 1716 1717 1718 1719 </span> 1720 1721 1722 1723 </a> 1724 </li> 1725 1726 1727 1728 1729 1730 1731 1732 1733 1734 1735 <li class="md-nav__item"> 1736 <a href="../../api/feat/utils/image_operations/" class="md-nav__link"> 1737 1738 1739 1740 <span class="md-ellipsis"> 1741 1742 1743 image_operations 1744 1745 1746 1747 </span> 1748 1749 1750 1751 </a> 1752 </li> 1753 1754 1755 1756 1757 1758 1759 1760 1761 1762 1763 <li class="md-nav__item"> 1764 <a href="../../api/feat/utils/io/" class="md-nav__link"> 1765 1766 1767 1768 <span class="md-ellipsis"> 1769 1770 1771 io 1772 1773 1774 1775 </span> 1776 1777 1778 1779 </a> 1780 </li> 1781 1782 1783 1784 1785 1786 1787 1788 1789 1790 1791 <li class="md-nav__item"> 1792 <a href="../../api/feat/utils/mp_plotting/" class="md-nav__link"> 1793 1794 1795 1796 <span class="md-ellipsis"> 1797 1798 1799 mp_plotting 1800 1801 1802 1803 </span> 1804 1805 1806 1807 </a> 1808 </li> 1809 1810 1811 1812 1813 1814 1815 1816 1817 1818 1819 <li class="md-nav__item"> 1820 <a href="../../api/feat/utils/muscle_to_landmark/" class="md-nav__link"> 1821 1822 1823 1824 <span class="md-ellipsis"> 1825 1826 1827 muscle_to_landmark 1828 1829 1830 1831 </span> 1832 1833 1834 1835 </a> 1836 </li> 1837 1838 1839 1840 1841 1842 1843 1844 1845 1846 1847 <li class="md-nav__item"> 1848 <a href="../../api/feat/utils/region_maps/" class="md-nav__link"> 1849 1850 1851 1852 <span class="md-ellipsis"> 1853 1854 1855 region_maps 1856 1857 1858 1859 </span> 1860 1861 1862 1863 </a> 1864 </li> 1865 1866 1867 1868 1869 1870 1871 1872 1873 1874 1875 <li class="md-nav__item"> 1876 <a href="../../api/feat/utils/stats/" class="md-nav__link"> 1877 1878 1879 1880 <span class="md-ellipsis"> 1881 1882 1883 stats 1884 1885 1886 1887 </span> 1888 1889 1890 1891 </a> 1892 </li> 1893 1894 1895 1896 1897 </ul> 1898 </nav> 1899 1900 </li> 1901 1902 1903 1904 1905 1906 1907 1908 1909 1910 1911 <li class="md-nav__item"> 1912 <a href="../../api/feat/transforms/" class="md-nav__link"> 1913 1914 1915 1916 <span class="md-ellipsis"> 1917 1918 1919 feat.transforms 1920 1921 1922 1923 </span> 1924 1925 1926 1927 </a> 1928 </li> 1929 1930 1931 1932 1933 </ul> 1934 </nav> 1935 1936 </li> 1937 1938 1939 1940 </ul> 1941</nav> 1942 </div> 1943 </div> 1944 </div> 1945 1946 1947 1948 <div class="md-sidebar md-sidebar--secondary" data-md-component="sidebar" data-md-type="toc" > 1949 <div class="md-sidebar__scrollwrap"> 1950 <div class="md-sidebar__inner"> 1951 1952 1953<nav class="md-nav md-nav--secondary" aria-label="Table of contents"> 1954 1955
1956 1957 1958 1959 1960 <label class="md-nav__title" for="__toc"> 1961 <span class="md-nav__icon md-icon"></span> 1962 Table of contents 1963 </label> 1964 <ul class="md-nav__list" data-md-component="toc" data-md-scrollfix> 1965 1966 <li class="md-nav__item"> 1967 <a href="#common-questions" class="md-nav__link"> 1968 <span class="md-ellipsis"> 1969 1970 Common questions 1971 1972 </span> 1973 </a> 1974 1975 <nav class="md-nav" aria-label="Common questions"> 1976 <ul class="md-nav__list"> 1977 1978 <li class="md-nav__item"> 1979 <a href="#py-feat-is-detecting-multiple-faces-in-an-image-with-a-single-face" class="md-nav__link"> 1980 <span class="md-ellipsis"> 1981 1982 Py-feat is detecting multiple faces in an image with a single face 1983 1984 </span> 1985 </a> 1986 1987</li> 1988 1989 <li class="md-nav__item"> 1990 <a href="#in-what-order-are-detected-faces-returned" class="md-nav__link"> 1991 <span class="md-ellipsis"> 1992 1993 In what order are detected faces returned? 1994 1995 </span> 1996 </a> 1997 1998</li> 1999 2000 <li class="md-nav__item"> 2001 <a href="#py-feat-is-treating-the-same-person-as-multiple-identities-or-treating-different-people-as-the-same-identity" class="md-nav__link"> 2002 <span class="md-ellipsis"> 2003 2004 Py-feat is treating the same person as multiple identities or treating different people as the same identity 2005 2006 </span> 2007 </a> 2008 2009</li> 2010 2011 <li class="md-nav__item"> 2012 <a href="#how-can-i-speed-things-up-and-control-memory-usage" class="md-nav__link"> 2013 <span class="md-ellipsis"> 2014 2015 How can I speed things up and control memory usage? 2016 2017 </span> 2018 </a> 2019 2020</li> 2021 2022 <li class="md-nav__item"> 2023 <a href="#how-fast-is-video-processing" class="md-nav__link"> 2024 <span class="md-ellipsis"> 2025 2026 How fast is video processing? 2027 2028 </span> 2029 </a> 2030 2031</li> 2032 2033 </ul> 2034 </nav> 2035 2036</li> 2037 2038 <li class="md-nav__item"> 2039 <a href="#known-issues" class="md-nav__link"> 2040 <span class="md-ellipsis"> 2041 2042 Known issues 2043 2044 </span> 2045 </a> 2046 2047</li> 2048 2049 <li class="md-nav__item"> 2050 <a href="#community-getting-help" class="md-nav__link"> 2051 <span class="md-ellipsis"> 2052 2053 Community & getting help 2054 2055 </span> 2056 </a> 2057 2058</li> 2059 2060 </ul> 2061 2062</nav> 2063 </div> 2064 </div> 2065 </div> 2066 2067 2068 2069 <div class="md-content" data-md-component="content"> 2070 2071 2072 2073 2074 2075 2076 2077 2078 <nav class="md-path" aria-label="Navigation" > 2079 <ol class="md-path__list"> 2080 2081 2082 2083 2084 <li class="md-path__item"> 2085 <a href="../.." class="md-path__link"> 2086 2087 <span class="md-ellipsis"> 2088 Home 2089 </span> 2090 2091 </a> 2092 </li> 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 <li class="md-path__item"> 2103 <a href="../installation/" class="md-path__link"> 2104 2105 <span class="md-ellipsis"> 2106 Getting Started 2107 </span> 2108 2109 </a> 2110 </li> 2111 2112 2113 2114 2115 </ol> 2116 </nav> 2117 2118 2119 <article class="md-content__inner md-typeset"> 2120 2121 2122 2123 2124 2125 2126 2127 2128<div class="marimo-book-buttons" data-placement="header"> 2129<a class="marimo-book-button marimo-book-button-github" href="https://github.com/cosanlab/py-feat/blob/main/docs/pages/faqs.md" target="_blank" rel="noopener" aria-label="View the source on GitHub" title="View the source on GitHub"><svg class="marimo-book-button-icon" viewBox="0 0 24 24" aria-hidden="true" focusable="false"><path d="M12 .3a12 12 0 0 0-3.79 23.4c.6.11.82-.26.82-.58v-2.05c-3.34.72-4.04-1.61-4.04-1.61-.55-1.39-1.34-1.76-1.34-1.76-1.08-.74.08-.73.08-.73 1.2.09 1.83 1.24 1.83 1.24 1.07 1.84 2.81 1.31 3.5 1 .11-.78.42-1.31.76-1.61-2.67-.3-5.47-1.33-5.47-5.93 0-1.31.46-2.38 1.24-3.22-.13-.31-.54-1.53.11-3.18 0 0 1.01-.32 3.31 1.23a11.5 11.5 0 0 1 6 0c2.31-1.55 3.31-1.23 3.31-1.23.66 1.65.25 2.87.13 3.18.77.84 1.24 1.91 1.24 3.22 0 4.61-2.81 5.62-5.49 5.92.42.36.81 1.1.81 2.22v3.29c0 .32.21.69.83.58A12 12 0 0 0 12 .3"/></svg><span class="marimo-book-button-label">View on GitHub</span></a> 2130</div> 2131
2132<h1 id="faqs-tips-known-issues">FAQs, Tips & Known Issues<a class="headerlink" href="#faqs-tips-known-issues" title="Permanent link">¶</a></h1> 2133<p>Here are a few general guidelines and common questions when using <code>py-feat</code>. We are always actively trying to improve the toolbox and make it easier to use â please help us out by contributing on GitHub!</p> 2134<div class="admonition note"> 2135<p class="admonition-title">Note</p> 2136<p><strong>Always spot-check your detections!</strong> While we've done our best to thoroughly test, benchmark, and document all the <a href="../models/">pre-trained models</a> included in Py-Feat, it's always possible that real-world images and videos reveal a quirk or limitation of an otherwise high-performing detector.</p> 2137</div> 2138<h2 id="common-questions">Common questions<a class="headerlink" href="#common-questions" title="Permanent link">¶</a></h2> 2139<h3 id="py-feat-is-detecting-multiple-faces-in-an-image-with-a-single-face">Py-feat is detecting multiple faces in an image with a single face<a class="headerlink" href="#py-feat-is-detecting-multiple-faces-in-an-image-with-a-single-face" title="Permanent link">¶</a></h3> 2140<p>This can happen sometimes, particularly for videos, as our current models don't use information from one frame to inform predictions for another frame. Try increasing the <code>face_detection_threshold</code> argument to <code>Detectorv1.detect()</code> from the default of <code>0.5</code> to something like <code>0.8</code> or <code>0.9</code>. This will make the detector more conservative in what it considers a face. You can also switch to the more accurate detector with <code>Detectorv1(face_model='retinaface')</code> (88.9% vs 55.5% WIDERFACE-Hard AP), or use <code>Detectorv2</code>, which also uses RetinaFace.</p> 2141<h3 id="in-what-order-are-detected-faces-returned">In what order are detected faces returned?<a class="headerlink" href="#in-what-order-are-detected-faces-returned" title="Permanent link">¶</a></h3> 2142<p>Within a frame, faces are returned in the order the face detector emits them â this order is <strong>not</strong> guaranteed to be stable across frames, and a given person will not necessarily occupy the same row position from frame to frame. To group detections by person across a video, use the identity labels (<code>Fex.identities</code>, computed via <code>Fex.compute_identities()</code>) rather than row position. If you need a deterministic per-frame order, sort each frame's rows yourself (e.g. by <code>FaceRectX</code>).</p> 2143<h3 id="py-feat-is-treating-the-same-person-as-multiple-identities-or-treating-different-people-as-the-same-identity">Py-feat is treating the same person as multiple identities or treating different people as the same identity<a class="headerlink" href="#py-feat-is-treating-the-same-person-as-multiple-identities-or-treating-different-people-as-the-same-identity" title="Permanent link">¶</a></h3> 2144<p>Similar to the previous issue, you can control the sensitivity of how confidently identity embeddings are retreated as different using the <code>face_identity_threshold</code> argument to <code>Detectorv1.detect()</code> from the default of <code>0.8</code> to something higher or lower. This will make the detector more conservative or liberal respectively, in how distinct identity embeddings have to be to be considered different people.</p> 2145<h3 id="how-can-i-speed-things-up-and-control-memory-usage">How can I speed things up and control memory usage?<a class="headerlink" href="#how-can-i-speed-things-up-and-control-memory-usage" title="Permanent link">¶</a></h3> 2146<p>By default all images or video frames are processed independently in batches of size 1 using your CPU. If you have a GPU, use the <code>device</code> argument when initializing a detector to make use of it â <code>Detectorv1(device='cuda')</code> for NVIDIA GPUs or <code>Detectorv1(device='mps')</code> for Apple Silicon. Both <code>Detectorv1</code> and <code>Detectorv2</code> support CUDA and MPS. To perform detections in parallel, increase the <code>batch_size</code> argument to <code>Detectorv1.detect()</code> from the default of 1. The largest batch size you can use without crashing your kernel is limited by the amount of VRAM available to your GPU (or RAM if you're using CPU).</p> 2147<p>In order to use batching you must either: 2148- use a video - where frames are all assumed to have the same dimensions 2149- use a list of images - where each image has the same dimensions 2150- use a list of images and set <code>output_size=(width, height)</code> in <code>.detect()</code> to resize all images to the same dimensions before processing</p> 2151<p>You can control parallelization of data loading using the <code>num_workers</code> argument to <code>.detect()</code>, which gets directly passed to pytorch's <a href="https://pytorch.org/tutorials/beginner/basics/data_tutorial.html">DataLoader</a>. Note that on Apple Silicon <code>num_workers > 0</code> is frequently <em>slower</em> than the default of <code>0</code>, so we recommend leaving it at <code>0</code> unless you've benchmarked a speedup on your own hardware.</p> 2152<h3 id="how-fast-is-video-processing">
2152How fast is video processing?<a class="headerlink" href="#how-fast-is-video-processing" title="Permanent link">¶</a></h3> 2153<p>Py-Feat's original release prioritized correctness and a seamless out-of-the-box experience over raw speed, and was quite slow (roughly 1â3 fps). That is no longer the case: <strong><code>Detectorv1</code> (v1) is now much faster, supports batching, and runs on the GPU (CUDA and MPS)</strong>, and <strong><code>Detectorv2</code> is faster still</strong> â it computes its whole multi-task output in a single forward pass. For current speed and accuracy numbers, see the <a href="../../benchmarks/accuracy/">accuracy</a> and <a href="../../benchmarks/Speed/">speed</a> benchmarks.</p> 2154<p>To get the most throughput, run on a GPU (<code>device='cuda'</code> or <code>device='mps'</code>) and increase <code>batch_size</code> in <code>.detect()</code> (see the question above).</p> 2155<p>By default, Py-feat avoids loading an entire video into memory, only loading frames as-needed, similar to Pytorch's <a href="https://pytorch.org/vision/main/auto_examples/others/plot_video_api.html#building-a-sample-read-video-function">unofficial video API</a>. This means you don't need to worry about your computer crashing if you're processing a video that doesn't fit into memory; the tradeoff is a small per-frame seek latency that grows with video length (later frames take slightly longer to load as the video is "seeked" to the correct time-point).</p> 2156<p>If you already know you have enough system memory to load the entire video at once, you can instead manually call <code>video_to_tensor('videofile.mp4')</code> from <code>feat.utils.io</code> and process the resulting tensor by passing <code>data_type='tensor'</code> to <code>Detectorv1.detect()</code>, batching as usual.</p> 2157<h2 id="known-issues">Known issues<a class="headerlink" href="#known-issues" title="Permanent link">¶</a></h2> 2158<ul> 2159<li>Detectors can be sensitive to differences in image sizes, such that very large or very small images result in very different predictions. This is largely due to differences in the hyper-parameters and image sizes used to train the underlying pre-trained models. To partially help with this issue, since <code>0.5.0</code> Py-Feat supports passing keyword arguments to underlying models during initialization or detection, e.g. <code>detector = Detectorv1(facepose_model_kwargs={'keep_top_k': 500})</code>. You can follow <a href="https://github.com/cosanlab/py-feat/issues/135">this issue</a> for more details.</li> 2160</ul> 2161<h2 id="community-getting-help">Community & getting help<a class="headerlink" href="#community-getting-help" title="Permanent link">¶</a></h2> 2162<ul> 2163<li><a href="https://www.askpbs.org/c/py-feat/26">Discourse Community</a>: a Stack Overflow-like forum where you can view, contribute, and vote on questions regarding <code>py-feat</code> usage. Please ask questions here first so other users can benefit from the answers!</li> 2164<li><a href="https://github.com/cosanlab/py-feat/issues">Open a GitHub issue</a>
2164 for all code-related problems. You can also do so by clicking the GitHub icon at the top of any page.</li> 2165</ul> 2166 2167 2168 2169 2170 2171 2172 2173 2174 2175 2176 2177 2178 2179 </article> 2180 </div> 2181 2182
2183<script>var target=document.getElementById(location.hash.slice(1));target&&target.name&&(target.checked=target.name.startsWith("__tabbed_"))</script>
2183 2184 </div> 2185 2186 </main> 2187 2188 <footer class="md-footer"> 2189 2190 2191 2192 <nav class="md-footer__inner md-grid" aria-label="Footer" > 2193 2194 2195 <a href="../au_reference/" class="md-footer__link md-footer__link--prev" aria-label="Previous: AU Reference"> 2196 <div class="md-footer__button md-icon"> 2197 2198 <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> 2199 </div> 2200 <div class="md-footer__title"> 2201 <span class="md-footer__direction"> 2202 Previous 2203 </span> 2204 <div class="md-ellipsis"> 2205 AU Reference 2206 </div> 2207 </div> 2208 </a> 2209 2210 2211 2212 <a href="../../basic_tutorials/Detecting_Images/" class="md-footer__link md-footer__link--next" aria-label="Next: Detecting Images"> 2213 <div class="md-footer__title"> 2214 <span class="md-footer__direction"> 2215 Next 2216 </span> 2217 <div class="md-ellipsis"> 2218 Detecting Images 2219 </div> 2220 </div> 2221 <div class="md-footer__button md-icon"> 2222 2223 <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> 2224 </div> 2225 </a> 2226 2227 </nav> 2228 2229 2230 <div class="md-footer-meta md-typeset"> 2231 <div class="md-footer-meta__inner md-grid"> 2232 <div class="md-copyright"> 2233 2234 <div class="md-copyright__highlight"> 2235 2022 · Made with <a href="https://marimobook.org" target="_blank" rel="noopener">Marimo-Book</a> 2236 </div> 2237 2238 2239</div> 2240 2241 </div> 2242 </div> 2243</footer> 2244 2245 </div> 2246 <div class="md-dialog" data-md-component="dialog"> 2247 <div class="md-dialog__inner md-typeset"></div> 2248 </div> 2249 2250 2251 2252 2253 2254
2254<script id="__config" type="application/json">{"annotate": null, "base": "../..", "features": ["navigation.sections", "navigation.expand", "navigation.path", "navigation.footer", "navigation.instant", "navigation.tracking", "search.suggest", "search.highlight", "search.share", "content.code.copy", "content.code.annotate", "toc.follow", "navigation.indexes"], "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>
2254 2255 2256 2257
2257<script src="../../assets/javascripts/bundle.d7400e89.min.js"></script>
2257 2258 2259
2259<script src="https://cdn.jsdelivr.net/npm/mathjax@3/es5/tex-mml-chtml.js"></script>
2259 2260 2261
2261<script src="../../javascripts/mathjax.js?v=0.1.26"></script>
2261 2262 2263
2263<script src="../../javascripts/marimo_book.js?v=0.1.26" defer></script>
2263 2264 2265 2266 </body> 2267</html>
Line numbers count LF bytes from the start of the resource, as the search results do. Vendor segments are library code the classifier recognised; they are stored but not indexed. Bytes are shown as Latin1 characters, one per byte.