1 2<html> 3<head> 4 <meta charset="utf-8"> 5<!-- Meta tags for social media banners, these should be filled in appropriately as they are your "business card" --> 6<!-- Replace the content tag with appropriate information --> 7<meta name="description" content="ATK: Automatic Task-driven Keypoint Selection for Robust Policy Learning. Learn how ATK improves robustness in visuomotor policy-learning through task-driven keypoint selection."> 8<meta property="og:title" content="ATK: Automatic Task-driven Keypoint Selection for Robust Policy Learning"/> 9<meta property="og:description" content="We purpose ATK, a novel method for robust policy learning using task-driven keypoint selection. Explore its applications in sim-to-real transfer and real-world imitation learning."/> 10<meta property="og:url" content="https://yourwebsite.com/ATK"/> 11 <!-- Path to banner image, should be in the path listed below. Optimal dimenssions are 1200X630--> 12 <meta property="og:image" content="static/image/shoe_img.png" /> 13 <meta property="og:image:width" content="1200"/> 14 <meta property="og:image:height" content="630"/> 15 16 17 <meta name="twitter:title" content="ATK: Automatic Task-driven Keypoint Selection for Robust Policy Learning"> 18 <meta name="twitter:description" content="We purpose ATK, a novel method for robust policy learning using task-driven keypoint selection. Explore its applications in sim-to-real transfer and real-world imitation learning."/> 19 <!-- Path to banner image, should be in the path listed below. Optimal dimenssions are 1200X600--> 20 <meta name="twitter:image" content="static/images/shoe_img.png"> 21 <meta name="twitter:card" content="summary_large_image"> 22 <!-- Keywords for your paper to be indexed by--> 23 <meta name="keywords" content="KEYWORDS SHOULD BE PLACED HERE"> 24 <meta name="viewport" content="width=device-width, initial-scale=1"> 25 26 <title>ATK-Policy</title> 27 28 <!-- Preload critical resources --> 29 <link rel="preload" href="https://fonts.googleapis.com/css2?family=Inter:wght@400;600;800&display=swap" as="style"> 30 <link rel="preload" href="https://cdnjs.cloudflare.com/ajax/libs/font-awesome/6.0.0/css/all.min.css" as="style"> 31 32 <link rel="icon" type="image/x-icon" href="static/images/logo.ico"> 33 <link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" 34 rel="stylesheet"> 35 36 <link rel="stylesheet" href="static/css/bulma.min.css"> 37 <link rel="stylesheet" href="static/css/bulma-carousel.min.css"> 38 <link rel="stylesheet" href="static/css/bulma-slider.min.css"> 39 <link rel="stylesheet" href="static/css/fontawesome.all.min.css"> 40 <link rel="stylesheet" 41 href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css"> 42 <link rel="stylesheet" href="static/css/index.css"> 43 44
44<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
44 45
45<script src="https://documentcloud.adobe.com/view-sdk/main.js"></script>
45 46
46<script defer src="static/js/fontawesome.all.min.js"></script>
46 47
47<script src="static/js/bulma-carousel.min.js"></script>
47 48
48<script src="static/js/bulma-slider.min.js"></script>
48 49
49<script src="static/js/index.js"></script>
49 50</head> 51<body> 52 <!-- Google Fonts --> 53<link href="https://fonts.googleapis.com/css2?family=Inter:wght@400;600;800&display=swap" rel="stylesheet"> 54 55<style> 56 @supports not (container-type: inline-size) { 57 /* Fallback for older browsers */ 58 .videos-row { display: flex; flex-wrap: wrap; } 59 .video-container { flex: 1; min-width: 300px; } 60 } 61</style> 62 63<section class="hero"> 64 <!-- Subtle texture overlay --> 65 <div class="texture-overlay"></div> 66 67 68 <!-- Top Text Section --> 69 <div class="hero-body top-row"> 70 <div class="container has-text-centered"> 71 <h1 class="title publication-title"> 72 <span class="atk-gradient">ATK</span>: Automatic Task-driven Keypoint Selection for Robust Policy Learning 73 </h1> 74 75 <div class="authors-size publication-authors"> 76 <div class="author-row"> 77 <span class="author-block"><a href="https://yunchuzhang.github.io/">Yunchu Zhang</a></span> 78 <span class="author-block"><a href="https://www.linkedin.com/in/shubham-mittal-445284213/">Shubham Mittal</a></span> 79 <span class="author-block"><a href="https://zoctipus.github.io/">Zhengyu Zhang</a></span> 80 <span class="author-block"><a href="https://kayke.xyz/">Liyiming Ke</a></span> 81 </div> 82 <div class="author-row"> 83 <span class="author-block"><a href="https://goodrobot.ai/">Siddhartha Srinivasa</a></span> 84 <span class="author-block"><a href="https://homes.cs.washington.edu/~abhgupta/">Abhishek Gupta</a></span> 85 </div> 86 </div> 87 88 <div class="is-size-5 publication-authors mt-2"> 89 <!-- <span class="author-block">University of Washington<br>CoRL 2025</span> --> 90 <div class="university-name is-size-7-mobile is-size-6-tablet is-size-5-desktop">University of Washington</div> 91 <!-- <div class="conference-name is-size-7-mobile is-size-6-tablet is-size-5-desktop">CoRL 2025</div> --> 92 </div> 93 94 <div class="publication-links mt-4"> 95 <span class="link-block"> 96 <a href="https://arxiv.org/pdf/2506.13867" target="_blank" 97 class="external-link"> 98 <span class="icon"><i class="fas fa-file-pdf"></i></span> 99 <span>Paper</span> 100 </a> 101 </span> 102 <span class="link-block"> 103 <a href="https://github.com/YOUR REPO HERE" target="_blank" 104 class="external-link"> 105 <span class="icon"><i class="fab fa-github"></i></span> 106 <span>Code (coming soon)</span> 107 </a> 108 </span> 109 </div> 110 </div> 111 </div> 112 113 <!-- Bottom Video Section --> 114 <div class="bottom-row"> 115 <div class="video-wrapper"> 116 <video loading="lazy" preload="none" muted loop playsinline> 117 <source src="static/videos/teaser.mp4" type="video/mp4"> 118 Your browser does not support the video tag. 119 </video> 120 </div> 121 </div> 122</section> 123 124<div class="all-content"> 125<div class="scale-wrapper"> 126<!-- Paper abstract --> 127 <section class="section is-light"> 128 <div class="container is-max-desktop"> 129 <div class="columns is-centered has-text-centered"> 130 <div class=" is-four-fifths"> 131 <h2 class="title-title">Motivation</h2> 132 <div class="content has-text-justified"> 133 <p> 134 135 Visuomotor policies often suffer from perceptual challenges, where visual differences between training and evaluation environments degrade policy performance. Policies relying on state estimations, like 6D pose, require task-specific tracking and are difficult to scale, while raw sensor-based policies may lack robustness to small visual disturbances. In this work, we leverage 2D keypoints â spatially consistent features in the image frame â as a flexible state representation for robust policy learning and apply it to both sim-to-real transfer and real-world imitation learning. However, the choice of which keypoints to use can vary across objects and tasks. We propose a novel method, <strong>ATK</strong>, to automatically select keypoints in a task-driven manner so that the chosen keypoints are predictive of optimal behavior for the given task. Our proposal optimizes for a minimal set of keypoints that focus on task-relevant parts while preserving policy performance and robustness. We distill expert data (either from an expert policy in simulation or a human expert) into a policy that operates on RGB images while tracking the selected keypoints. By leveraging pre-trained visual modules, our system effectively encodes states and transfers policies to the real-world evaluation scenario despite wide scene variations and perceptual challenges such as transparent objects, fine-grained tasks, and deformable objects manipulation. 136 </p> 137 138 139 </div> 140 </div> 141 </div> 142 </div> 143 </section> 144 <!-- End paper abstract --> 145 146<section class="teaser"> 147 <div class="container is-max-desktop"> 148 <div class="hero-body-method"> 149 <video poster="static/images/method_placeholder.png" id="tree" controls preload="metadata" class="manual-control" muted height="100%"> 150 <source src="static/videos/method.mp4" type="video/mp4"> 151 </video> 152 <div class="method-play-button" id="method-play-btn"> 153 <svg viewBox="0 0 24 24"> 154 <path d="M8 5v14l11-7z"/> 155 </svg> 156 </div> 157 </div> 158 </div> 159</section> 160 161<section class="section is-light global-header"> 162 <div class="container"> 163 <h2 class="title-title has-text-centered" style="color: rgb(0, 0, 0);">Our Method</h2> 164 <!-- Results Section --> 165<section class="section is-light result-container result-section"> 166 <div class="container"> 167 <div class="results"> 168 <div class="result is-centered"> 169 <!-- Left Image --> 170 <div class=""> 171 <img src="static/images/method.jpg" alt="Method" style="width: 100%; height: auto;"> 172 </div> 173 </div>
174 <figcaption> 175 176 <p> <strong>ATK automatically selects minimal yet necessary information for task execution</strong> by distilling 177 expert data (either from an expert policy in simulation or a human expert) into a policy that operates on a 178 selective subset of keypoints and optimizing the selection mask. Once the keypoints are identified, they are 179 transferred from the training set to the real-world evaluation scenario. Finally, the keypoint-based policy is 180 transferred to the evaluation scenario, taking as input RGB images while tracking the transferred keypoints</p> 181 </figcaption> 182 </div> 183 </div> 184</section> 185 186</section> 187<!-- Experiments --> 188<!-- Immitation Learning --> 189 190<!-- Global Header --> 191<section class="section is-light global-header"> 192 <div class="container"> 193 <h2 class="title-title has-text-centered" style="color: rgb(0, 0, 0);">Experiments</h2> 194 195 <h2 class="title has-text-centered" style="color: rgb(0, 0, 0);">1. Immitation Learning Robustness</h2> 196 <p class="subtitle explain has-text-centered" >ATK shows robustness to random background, distractors and light</p> 197 198</section> 199<section class="columns-tasks section is-dark category-section" > 200<div class="sticky-container"> 201 <div class="sticky-header"> 202 <div class="button-row"> 203 <button class="button task-button is-active" data-task="towel-folding">Towel Folding</button> 204 <button class="button task-button" data-task="hanging-blanket">Hanging Blanket</button> 205 <button class="button task-button" data-task="grape-oven">Grape-Oven</button> 206 <button class="button task-button" data-task="pan-filling">Pan Filling</button> 207 </div> 208 <!-- <div class="tabs is-centered is-boxed sticky-toggle-bar"> 209 <ul> 210 <li data-view="side-views" class="is-active"> 211 <a>Global View</a> 212 </li> 213 <li data-view="robot"> 214 <a>Robot Camera View</a> 215 </li> 216 </ul> 217 </div> --> 218 </div> 219 <!-- </div> --> 220 <div class="container video-gallery-wrapper" id="towel-folding"> 221 <!--------------------------------------------- Towel Folding -------------------------------------------------------> 222 223 <!-- Video Gallery Section --> 224 <section class="section is-light video-gallery" > 225 226 <!-- Task Description --> 227 <!-- <h4 class="is-5 task-title";> 228 <strong>Towel Folding</strong> 229 </h4> --> 230 231 <!-- Combined Video Demonstration --> 232 <div id="task-teaser"> 233 <!-- <div class="video-container"> --> 234 <!-- <video loading="lazy" preload="none" muted loop playsinline > 235 <source src="static/videos/kp_filtering/towel_combined.mp4" type="video/mp4"> 236 Your browser does not support the video tag. 237 </video> --> 238 <!-- </div> --> 239 <div class="teaser-row"> 240 <!-- Left Video --> 241 <div class="video-container"> 242 <h4 class="text-teaser";> 243 Automatic Task-driven Keypoints Filteringâ </h4> 244 <video loading="lazy" preload="none" muted loop playsinline > 245 <source src="static/videos/teaser/towel_filtering.mp4" type="video/mp4"> 246 Your browser does not support the video tag. 247 </video> 248 </div> 249 250 <!-- Center Image --> 251 <div class="video-container"> 252 <h4 class="text-teaser";> 253 Keypoints Transfer during Inferenceââ </h4> 254 <div class="image-grid"> 255 <div class="image-container"> 256 <img src="static/images/towel_2.png" alt="Image 1" class="teaser-image"> 257 </div> 258 <div class="image-container"> 259 <img src="static/images/towel_3.png" alt="Image 2" class="teaser-image"> 260 </div> 261 <div class="image-container"> 262 <img src="static/images/towel_4.png" alt="Image 3" class="teaser-image"> 263 </div> 264 <div class="image-container"> 265 <img src="static/images/towel_5.png" alt="Image 4" class="teaser-image"> 266 </div> 267 </div> 268 </div> 269 270 <!-- Right Video --> 271 <div> 272 <h4 class="text-teaser";> 273 Keypoints Tracking during Rolloutâ </h4> 274 <div class="video-container"> 275 <div class="vid-tex-overlay"> 276 <video loading="lazy" preload="none" muted loop playsinline > 277 <source src="static/videos/teaser/towel_tracking.mp4" type="video/mp4"> 278 Your browser does not support the video tag. 279 </video> 280 <img src="static/images/8x_black.svg" class="eight-teaser" > 281 </div> 282 </div> 283 </div> 284 </div> 285 </div> 286 287 288 <div class="container four-condn-container"> 289 290 <!-- Video Content Container --> 291 <div class="video-content"> 292 <!-- Side Views (DJI & Moving) --> 293 <div class="view-content" id="side-views"> 294 <div class="side-views-container"> 295 <!-- DJI Camera View | Moving View--> 296 <div id="dji-view-towel-folding"> 297 <!-- Task 1 --> 298 <div class="task-section"> 299 <!-- Row 1 --> 300 301 302 <!-- Task 1: Towel Folding --> 303 <section class="section task-section"> 304 <div class="condition-section"> 305 <div class="video-selector-wrapper"> 306 <div class="video-selector-container"> 307 <div class="select mb-2"> 308 <select class="video-selector" data-condition="random-position" data-view="dji"> 309 <option value="1">Rollout 1</option> 310 <option value="2">Rollout 2</option> 311 <option value="3">Rollout 3</option> 312 <option value="4">Rollout 4</option> 313 314 </select> 315 316 317 </div> 318 319 </div> 320 <div class="video-selector-condn"> 321 <h5 class="subtitle is-5">Random Position</h5> 322 </div> 323 324 </div> 325 326 <div class="videos-row"> 327 <!-- First Row: Two Videos --> 328 <div class="video-container"> 329 <div class="text-atk"> 330 <p>ATK (static view)</p> 331 </div> 332 <div class="vid-tex-overlay"> 333 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 334 <source src="static/videos/dji/towel/rp1.mp4" type="video/mp4"> 335 Your browser does not support the video tag. 336 </video> 337 <img src="static/images/6x_black.svg" class="six" > 338 </div> 339 340 </div> 341 <div class="video-container"> 342 <div class="text-atk"> 343 <p>ATK (moving view)</p> 344 </div> 345 <div class="vid-tex-overlay"> 346 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 347 <source src="static/videos/iphone/towel/rp1.mp4" type="video/mp4"> 348 Your browser does not support the video tag. 349 </video> 350 <img src="static/images/6x_black.svg" class="six" > 351 </div> 352 </div> 353 </div> 354 355 356 <div class="videos-row-baselines"> 357 <!-- Second Row: Three Videos --> 358 <div class="video-container success"> 359 <div class="text-baseline"> 360 <p>R3M</p> 361 </div> 362 <div class="vid-tex-overlay"> 363 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 364 <source src="static/videos/baselines/towel/r3m/org.mp4" type="video/mp4"> 365 Your browser does not support the video tag. 366 </video> 367 <img src="static/images/tick.svg" class="tick-mark" > 368 <img src="static/images/6x_black.svg" class="six_baseline" > 369 </div> 370 371 </div> 372 373 <div class="video-container"> 374 <div class="text-baseline"> 375 <p>Depth</p> 376 </div> 377 <div class="vid-tex-overlay"> 378 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 379 <source src="static/videos/baselines/towel/depth/org.mp4" type="video/mp4"> 380 Your browser does not support the video tag. 381 </video> 382 <img src="static/images/cross.svg" class="tick-mark" > 383 <img src="static/images/6x_black.svg" class="six_baseline" > 384 </div> 385 386 </div> 387 <div class="video-container"> 388 <div class="text-baseline"> 389 <p>Point Cloud</p> 390 </div> 391 <div class="vid-tex-overlay"> 392 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 393 <source src="static/videos/baselines/towel/pointcloud/org.mp4" type="video/mp4"> 394 Your browser does not support the video tag. 395 </video> 396 <img src="static/images/cross.svg" class="tick-mark" > 397 <img src="static/images/6x_black.svg" class="six_baseline" > 398 </div> 399 400 </div> 401 </div> 402 </div> 403 <div class="condition-section"> 404 <div class="video-selector-wrapper"> 405 <div class="video-selector-container"> 406 <div class="select mb-2"> 407 <select class="video-selector" data-condition="random-distractors" data-view="dji"> 408 <option value="1">Rollout 1</option> 409 <option value="2">Rollout 2</option> 410 <option value="3">Rollout 3</option> 411 <option value="4">Rollout 4</option> 412 </select> 413 </div> 414 415 </div> 416 <div class="video-selector-condn"> 417 <h5 class="subtitle is-5">Random Distractors</h5> 418 </div> 419 420 </div> 421 422 <div class="videos-row"> 423 <!-- First Row: Two Videos --> 424 <div class="video-container"> 425 <div class="text-atk"> 426 <p>ATK (static view)</p> 427 </div> 428 <div class="vid-tex-overlay"> 429 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 430 <source src="static/videos/dji/towel/rd1.mp4" type="video/mp4"> 431 Your browser does not support the video tag. 432 </video> 433 <img src="static/images/6x_black.svg" class="six" > 434
434 </div> 435 436 </div> 437 <div class="video-container"> 438 <div class="text-atk"> 439 <p>ATK (moving view)</p> 440 </div> 441 <div class="vid-tex-overlay"> 442 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 443 <source src="static/videos/iphone/towel/rd1.mp4" type="video/mp4"> 444 Your browser does not support the video tag. 445 </video> 446 <img src="static/images/6x_black.svg" class="six" > 447 </div> 448 </div> 449 </div> 450 451 452 <div class="videos-row-baselines"> 453 <!-- Second Row: Three Videos --> 454 <div class="video-container success"> 455 <div class="text-baseline"> 456 <p>R3M</p> 457 </div> 458 <div class="vid-tex-overlay"> 459 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 460 <source src="static/videos/baselines/towel/r3m/dist.mp4" type="video/mp4"> 461 Your browser does not support the video tag. 462 </video> 463 <img src="static/images/cross.svg" class="tick-mark" > 464 <img src="static/images/6x_black.svg" class="six_baseline" > 465 466 </div> 467 468 </div> 469 470 <div class="video-container"> 471 <div class="text-baseline"> 472 <p>Depth</p> 473 </div> 474 <div class="vid-tex-overlay"> 475 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 476 <source src="static/videos/baselines/towel/depth/dist.mp4" type="video/mp4"> 477 Your browser does not support the video tag. 478 </video> 479 <img src="static/images/6x_black.svg" class="six_baseline" > 480 <img src="static/images/cross.svg" class="tick-mark" > 481 </div> 482 483 </div> 484 <div class="video-container"> 485 <div class="text-baseline"> 486 <p>Point Cloud</p> 487 </div> 488 <div class="vid-tex-overlay"> 489 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 490 <source src="static/videos/baselines/towel/pointcloud/dist.mp4" type="video/mp4"> 491 Your browser does not support the video tag. 492 </video> 493 <img src="static/images/6x_black.svg" class="six_baseline" > 494 <img src="static/images/cross.svg" class="tick-mark" > 495 </div> 496 497 </div> 498 </div> 499 </div> 500 <div class="condition-section"> 501 <div class="video-selector-wrapper"> 502 <div class="video-selector-container"> 503 <div class="select mb-2"> 504 <select class="video-selector" data-condition="random-backgrounds" data-view="dji"> 505 <option value="1">Rollout 1</option> 506 <option value="2">Rollout 2</option> 507 <option value="3">Rollout 3</option> 508 <option value="4">Rollout 4</option> 509 </select> 510 </div> 511 512 </div> 513 <div class="video-selector-condn"> 514 <h5 class="subtitle is-5">
514Random Backgrounds</h5> 515 </div> 516 517 </div> 518 519 <div class="videos-row"> 520 <!-- First Row: Two Videos --> 521 <div class="video-container"> 522 <div class="text-atk"> 523 <p>ATK (static view)</p> 524 </div> 525 <div class="vid-tex-overlay"> 526 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 527 <source src="static/videos/dji/towel/rb1.mp4" type="video/mp4"> 528 Your browser does not support the video tag. 529 </video> 530 <img src="static/images/6x_black.svg" class="six" > 531 </div> 532 533 </div> 534 <div class="video-container"> 535 <div class="text-atk"> 536 <p>ATK (moving view)</p> 537 </div> 538 <div class="vid-tex-overlay"> 539 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 540 <source src="static/videos/iphone/towel/rb1.mp4" type="video/mp4"> 541 Your browser does not support the video tag. 542 </video> 543 <img src="static/images/6x_black.svg" class="six" > 544 </div> 545 </div> 546 </div> 547 548 549 <div class="videos-row-baselines"> 550 <!-- Second Row: Three Videos --> 551 <div class="video-container success"> 552 <div class="text-baseline"> 553 <p>R3M</p> 554 </div> 555 <div class="vid-tex-overlay"> 556 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 557 <source src="static/videos/baselines/towel/r3m/bck.mp4" type="video/mp4"> 558 Your browser does not support the video tag. 559 </video> 560 <img src="static/images/6x_black.svg" class="six_baseline" > 561 <img src="static/images/cross.svg" class="tick-mark" > 562 </div> 563 564 </div> 565 566 <div class="video-container"> 567 <div class="text-baseline"> 568 <p>Depth</p> 569 </div> 570 <div class="vid-tex-overlay"> 571 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 572 <source src="static/videos/baselines/towel/depth/bck.mp4" type="video/mp4"> 573 Your browser does not support the video tag. 574 </video> 575 <img src="static/images/6x_black.svg" class="six_baseline" > 576 <img src="static/images/cross.svg" class="tick-mark" > 577 </div> 578 579 </div> 580 <div class="video-container"> 581 <div class="text-baseline"> 582 <p>Point Cloud</p> 583 </div> 584 <div class="vid-tex-overlay"> 585 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 586 <source src="static/videos/baselines/towel/pointcloud/bck.mp4" type="video/mp4"> 587 Your browser does not support the video tag. 588 </video> 589 <img src="static/images/6x_black.svg" class="six_baseline" > 590 <img src="static/images/cross.svg" class="tick-mark" > 591 </div> 592 593 </div> 594 </div> 595 </div> 596 597 <div class="condition-section"> 598 <div class="video-selector-wrapper"> 599 <div class="video-selector-container"> 600 <div class="select mb-2"> 601 <select class="video-selector" data-condition="random-lights" data-view="dji"> 602 <option value="1">Rollout 1</option> 603 <option value="2">Rollout 2</option> 604 </select> 605 </div> 606 607 </div> 608 <div class="video-selector-condn"> 609 <h5 class="subtitle is-5">Random Lights</h5> 610 </div> 611 612 </div> 613 614 <div class="videos-row"> 615 <!-- First Row: Two Videos --> 616 <div class="video-container"> 617 <div class="text-atk"> 618 <p>ATK (static view)</p> 619 </div> 620 <div class="vid-tex-overlay"> 621 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 622 <source src="static/videos/dji/towel/disco.mp4" type="video/mp4"> 623 Your browser does not support the video tag. 624 </video> 625 <img src="static/images/6x_black.svg" class="six" > 626 </div> 627 628 </div> 629 <div class="video-container"> 630 <div class="text-atk"> 631 <p>ATK (moving view)</p> 632 </div> 633 <div class="vid-tex-overlay"> 634 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 635 <source src="static/videos/iphone/towel/disco.mp4" type="video/mp4"> 636 Your browser does not support the video tag. 637 </video> 638 <img src="static/images/6x_black.svg" class="six" > 639 </div> 640 </div> 641 </div> 642 643 644 <div class="videos-row-baselines"> 645 <!-- Second Row: Three Videos --> 646 <div class="video-container success"> 647 <div class="text-baseline"> 648 <p>R3M</p> 649 </div> 650 <div class="vid-tex-overlay"> 651 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 652 <source src="static/videos/baselines/towel/r3m/disco.mp4" type="video/mp4"> 653 Your browser does not support the video tag. 654 </video> 655 <img src="static/images/6x_black.svg" class="six_baseline" > 656 <img src="static/images/cross.svg" class="tick-mark" > 657 </div> 658 659 </div> 660 661 <div class="video-container"> 662 <div class="text-baseline"> 663 <p>Depth</p> 664 </div> 665 <div class="vid-tex-overlay"> 666 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 667 <source src="static/videos/baselines/towel/depth/disco.mp4" type="video/mp4"> 668 Your browser does not support the video tag. 669 </video> 670 <img src="static/images/6x_black.svg" class="six_baseline" > 671 <img src="static/images/cross.svg" class="tick-mark" > 672 </div> 673 674 </div> 675 <div class="video-container"> 676 <div class="text-baseline"> 677 <p>Point Cloud</p> 678 </div> 679 <div class="vid-tex-overlay"> 680 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 681 <source src="static/videos/baselines/towel/pointcloud/disco.mp4" type="video/mp4"> 682 Your browser does not support the video tag. 683 </video> 684 <img src="static/images/6x_black.svg" class="six_baseline" > 685 <img src="static/images/cross.svg" class="tick-mark" > 686 </div> 687 </div> 688 </div> 689 </div> 690 </section> 691 </div> 692 </div> 693 </div> 694 </div> 695 </div> 696 </section> 697 </div> 698 699 <div class="container video-gallery-wrapper is-hidden" id="hanging-blanket"> 700 <!--------------------------------------------- Hanging Blanket -------------------------------------------------------> 701 702 <!-- Video Gallery Section --> 703 <section class="section is-light video-gallery" > 704 705 <!-- Task Description --> 706 <!-- <h4 class="is-5 task-title";> 707 <strong>Towel Folding</strong> 708 </h4> --> 709 710 <!-- Combined Video Demonstration --> 711 <div id="task-teaser"> 712 <!-- <div class="video-container"> --> 713 <!-- <video loading="lazy" preload="none" muted loop playsinline > 714 <source src="static/videos/kp_filtering/towel_c
714ombined.mp4" type="video/mp4"> 715 Your browser does not support the video tag. 716 </video> --> 717 <!-- </div> --> 718 <div class="teaser-row"> 719 <!-- Left Video --> 720 <div class="video-container"> 721 <h4 class="text-teaser";> 722 Automatic Task-driven Keypoints Filteringâ </h4> 723 <video loading="lazy" preload="none" muted loop playsinline > 724 <source src="static/videos/teaser/hanging_filtering.mp4" type="video/mp4"> 725 Your browser does not support the video tag. 726 </video> 727 </div> 728 729 <!-- Center Image --> 730 <div class="video-container"> 731 <h4 class="text-teaser";> 732 Keypoints Transfer during Inferenceââ </h4> 733 <div class="image-grid"> 734 <div class="image-container"> 735 <img src="static/images/hanging_2.png" alt="Image 1" class="teaser-image"> 736 </div> 737 <div class="image-container"> 738 <img src="static/images/hanging_3.png" alt="Image 2" class="teaser-image"> 739 </div> 740 <div class="image-container"> 741 <img src="static/images/hanging_4.png" alt="Image 3" class="teaser-image"> 742 </div> 743 <div class="image-container"> 744 <img src="static/images/hanging_5.png" alt="Image 4" class="teaser-image"> 745 </div> 746 </div> 747 </div> 748 749 <!-- Right Video --> 750 <div> 751 <h4 class="text-teaser";> 752 Keypoints Tracking during Rolloutâ </h4> 753 <div class="video-container"> 754 <div class="vid-tex-overlay"> 755 <video loading="lazy" preload="none" muted loop playsinline > 756 <source src="static/videos/teaser/hanging_tracking.mp4" type="video/mp4"> 757 Your browser does not support the video tag. 758 </video> 759 <img src="static/images/8x_black.svg" class="eight-teaser" > 760 </div> 761 </div> 762 </div> 763 </div> 764 </div> 765 766 767 <div class="container four-condn-container"> 768 769 <!-- Video Content Container --> 770 <div class="video-content"> 771 <!-- Side Views (DJI & Moving) --> 772 <div class="view-content" id="side-views"> 773 <div class="side-views-container"> 774 <!-- DJI Camera View | Moving View--> 775 <div id="dji-view-hanging-blanket"> 776 <!-- Task 1 --> 777 <div class="task-section"> 778 <!-- Row 1 --> 779 780 781 782 <section class="section task-section"> 783 <div class="condition-section"> 784 <div class="video-selector-wrapper"> 785 <div class="video-selector-container"> 786 <div class="select mb-2"> 787 <select class="video-selector" data-condition="random-position" data-view="dji"> 788 <option value="1">Rollout 1</option> 789 <option value="2">Rollout 2</option> 790 <option value="3">Rollout 3</option> 791 <option value="4">Rollout 4</option> 792 793 </select> 794 795 796 </div> 797 798 </div> 799 <div class="video-selector-condn"> 800 <h5 class="subtitle is-5">Random Position</h5> 801 </div> 802 803 </div> 804 805 <div class="videos-row"> 806 <!-- First Row: Two Videos --> 807 <div class="video-container"> 808 <div class="text-atk"> 809 <p>ATK (static view)</p> 810 </div> 811 <div class="vid-tex-overlay"> 812 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 813 <source src="static/videos/dji/hanging/rp1.mp4" type="video/mp4"> 814 Your browser does not support the video tag. 815 </video> 816 <img src="static/images/6x_black.svg" class="six" > 817 </div> 818 819 </div> 820 <div class="video-container"> 821 <div class="text-atk"> 822 <p>ATK (moving view)</p> 823 </div> 824 <div class="vid-tex-overlay"> 825 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 826 <source src="static/videos/iphone/hanging/rp1.mp4" type="video/mp4"> 827 Your browser does not support the video tag. 828 </video> 829 <img src="static/images/6x_black.svg" class="six" > 830 </div> 831 </div> 832 </div> 833 834 835 <div class="videos-row-baselines"> 836 <!-- Second Row: Three Videos --> 837 <div class="video-container success"> 838 <div class="text-baseline"> 839 <p>R3M</p> 840 </div> 841 <div class="vid-tex-overlay"> 842 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 843 <source src="static/videos/baselines/hanging/r3m/org.mp4" type="video/mp4"> 844 Your browser does not support the video tag. 845 </video> 846 <img src="static/images/tick.svg" class="tick-mark" > 847 <img src="static/images/6x_black.svg" class="six_baseline" > 848 </div> 849 850 </div> 851 852 <div class="video-container"> 853 <div class="text-baseline"> 854 <p>Depth</p> 855 </div> 856 <div class="vid-tex-overlay"> 857 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 858 <source src="static/videos/baselines/hanging/depth/org.mp4" type="video/mp4"> 859 Your browser does not support the video tag. 860 </video> 861 <img src="static/images/cross.svg" class="tick-mark" > 862 <img src="static/images/6x_black.svg" class="six_baseline" > 863 </div> 864 865 </div> 866 <div class="video-container"> 867 <div class="text-baseline"> 868 <p>Point Cloud</p> 869 </div> 870 <div class="vid-tex-overlay"> 871 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 872 <source src="static/videos/baselines/hanging/pointcloud/org.mp4" type="video/mp4"> 873 Your browser does not support the video tag. 874 </video> 875 <img src="static/images/cross.svg" class="tick-mark" > 876 <img src="static/images/6x_black.svg" class="six_baseline" > 877 </div> 878 879 </div> 880 </div> 881 </div> 882 <div class="condition-section"> 883 <div class="video-selector-wrapper"> 884 <div class="video-selector-container"> 885 <div class="select mb-2"> 886 <select class="video-selector" data-condition="random-distractors" data-view="dji"> 887 <option value="1">Rollout 1</option> 888 <option value="2">Rollout 2</option> 889 <option value="3">Rollout 3</option> 890 <option value="4">Rollout 4</option> 891 </select> 892 </div> 893 894 </div> 895 <div class="video-selector-condn"> 896 <h5 class="subtitle is-5">Random Distractors</h5> 897 </div> 898 899 </div> 900 901 <div class="videos-row"> 902 <!-- First Row: Two Videos --> 903 <div class="video-container"> 904 <div class="text-atk"> 905 <p>ATK (static view)</p> 906 </div> 907 <div class="vid-tex-overlay"> 908 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 909 <source src="static/videos/dji/hanging/rd1.mp4" type="video/mp4"> 910 Your browser does not support the video tag. 911 </video> 912 <img src="static/images/6x_black.svg" class="six" > 913 </div> 914 915 </div> 916 <div class="video-container"> 917 <div class="text-atk"> 918 <p>ATK (moving view)</p> 919 </div> 920 <div class="vid-tex-overlay"> 921 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 922 <source src="static/videos/iphone/hanging/rd1.mp4" type="video/mp4"> 923 Your browser does not support the video tag. 924 </video> 925 <img src="static/images/6x_black.svg" class="six" > 926 </div> 927 </div> 928 </div> 929 930 931 <div class="videos-row-baselines"> 932 <!-- Second Row: Three Videos --> 933 <div class="video-container success"> 934 <div class="text-baseline"> 935 <p>R3M</p> 936 </div> 937 <div class="vid-tex-overlay"> 938 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 939 <source src="static/videos/baselines/hanging/r3m/dist.mp4" type="video/mp4"> 940 Your browser does not support the video tag. 941 </video> 942 <img src="static/images/cross.svg" class="tick-mark" > 943 <img src="static/images/6x_black.svg" class="six_baseline" > 944 945 </div> 946 947 </div> 948 949 <div class="video-container"> 950 <div class="text-baseline"> 951 <p>Depth</p> 952 </div> 953 <div class="vid-tex-overlay"> 954
954 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 955 <source src="static/videos/baselines/hanging/depth/dist.mp4" type="video/mp4"> 956 Your browser does not support the video tag. 957 </video> 958 <img src="static/images/6x_black.svg" class="six_baseline" > 959 <img src="static/images/cross.svg" class="tick-mark" > 960 </div> 961 962 </div> 963 <div class="video-container"> 964 <div class="text-baseline"> 965 <p>Point Cloud</p> 966 </div> 967 <div class="vid-tex-overlay"> 968 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 969 <source src="static/videos/baselines/hanging/pointcloud/dist.mp4" type="video/mp4"> 970 Your browser does not support the video tag. 971 </video> 972 <img src="static/images/6x_black.svg" class="six_baseline" > 973 <img src="static/images/cross.svg" class="tick-mark" > 974 </div> 975 976 </div> 977 </div> 978 </div> 979 <div class="condition-section"> 980 <div class="video-selector-wrapper"> 981 <div class="video-selector-container"> 982 <div class="select mb-2"> 983 <select class="video-selector" data-condition="random-backgrounds" data-view="dji"> 984 <option value="1">Rollout 1</option> 985 <option value="2">Rollout 2</option> 986 <option value="3">Rollout 3</option> 987 <option value="4">Rollout 4</option> 988 </select> 989 </div> 990 991 </div> 992 <div class="video-selector-condn"> 993 <h5 class="subtitle is-5">Random Backgrounds</h5> 994 </div> 995 996 </div> 997 998 <div class="videos-row"> 999 <!-- First Row: Two Videos --> 1000 <div class="video-container"> 1001 <div class="text-atk"> 1002 <p>ATK (static view)</p> 1003 </div> 1004 <div class="vid-tex-overlay"> 1005 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1006 <source src="static/videos/dji/hanging/rb1.mp4" type="video/mp4"> 1007 Your browser does not support the video tag. 1008 </video> 1009 <img src="static/images/6x_black.svg" class="six" > 1010 </div> 1011 1012 </div> 1013 <div class="video-container"> 1014 <div class="text-atk"> 1015 <p>ATK (moving view)</p> 1016 </div> 1017 <div class="vid-tex-overlay"> 1018 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1019 <source src="static/videos/iphone/hanging/rb1.mp4" type="video/mp4"> 1020 Your browser does not support the video tag. 1021 </video> 1022 <img src="static/images/6x_black.svg" class="six" > 1023 </div> 1024 </div> 1025 </div> 1026 1027 1028 <div class="videos-row-baselines"> 1029 <!-- Second Row: Three Videos --> 1030 <div class="video-container success"> 1031 <div class="text-baseline"> 1032 <p>R3M</p> 1033 </div> 1034 <div class="vid-tex-overlay"> 1035 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1036 <source src="static/videos/baselines/hanging/r3m/bck.mp4" type="video/mp4"> 1037 Your browser does not support the video tag. 1038 </video> 1039 <img src="static/images/6x_black.svg" class="six_baseline" > 1040 <img src="static/images/cross.svg" class="tick-mark" > 1041 </div> 1042 1043 </div> 1044 1045 <div class="video-container"> 1046 <div class="text-baseline"> 1047 <p>Depth</p> 1048 </div> 1049 <div class="vid-tex-overlay"> 1050 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1051 <source src="static/videos/baselines/hanging/depth/bck.mp4" type="video/mp4"> 1052 Your browser does not support the video tag. 1053 </video> 1054 <img src="static/images/6x_black.svg" class="six_baseline" > 1055 <img src="static/images/cross.svg" class="tick-mark" > 1056 </div> 1057 1058 </div> 1059 <div class="video-container"> 1060 <div class="text-baseline"> 1061 <p>Point Cloud</p> 1062 </div> 1063 <div class="vid-tex-overlay"> 1064 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1065 <source src="static/videos/baselines/hanging/pointcloud/bck.mp4" type="video/mp4"> 1066 Your browser does not support the video tag. 1067 </video> 1068 <img src="static/images/6x_black.svg" class="six_baseline" > 1069 <img src="static/images/cross.svg" class="tick-mark" > 1070 </div> 1071 1072 </div> 1073 </div> 1074 </div> 1075 1076 <div class="condition-section"> 1077 <div class="video-selector-wrapper"> 1078 <div class="video-selector-container"> 1079 <div class="select mb-2"> 1080 <select class="video-selector" data-condition="random-lights" data-view="dji"> 1081 <option value="1">Rollout 1</option> 1082 <option value="2">Rollout 2</option> 1083 </select> 1084 </div> 1085 1086 </div> 1087 <div class="video-selector-condn"> 1088 <h5 class="subtitle is-5">Random Lights</h5> 1089 </div> 1090 1091 </div> 1092 1093 <div class="videos-row"> 1094 <!-- First Row: Two Videos --> 1095 <div class="video-container"> 1096 <div class="text-atk"> 1097 <p>ATK (static view)</p> 1098 </div> 1099 <div class="vid-tex-overlay"> 1100 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1101 <source src="static/videos/dji/hanging/disco.mp4" type="video/mp4"> 1102 Your browser does not support the video tag. 1103 </video> 1104 <img src="static/images/6x_black.svg" class="six" > 1105 </div> 1106 1107 </div> 1108 <div class="video-container"> 1109 <div class="text-atk"> 1110 <p>ATK (moving view)</p> 1111 </div> 1112 <div class="vid-tex-overlay"> 1113 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1114 <source src="static/videos/iphone/hanging/disco.mp4" type="video/mp4"> 1115 Your browser does not support the video tag. 1116 </video> 1117 <img src="static/images/6x_black.svg" class="six" > 1118 </div> 1119 </div> 1120 </div> 1121 1122 1123 <div class="videos-row-baselines"> 1124 <!-- Second Row: Three Videos --> 1125 <div class="video-container success"> 1126 <div class="text-baseline"> 1127 <p>R3M</p> 1128 </div> 1129 <div class="vid-tex-overlay"> 1130 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1131 <source src="static/videos/baselines/hanging/r3m/disco.mp4" type="video/mp4"> 1132 Your browser does not support the video tag. 1133 </video> 1134 <img src="static/images/6x_black.svg" class="six_baseline" > 1135 <img src="static/images/cross.svg" class="tick-mark" > 1136 </div> 1137 1138 </div> 1139 1140 <div class="video-container"> 1141 <div class="text-baseline"> 1142 <p>Depth</p> 1143 </div> 1144 <div class="vid-tex-overlay"> 1145 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1146 <source src="static/videos/baselines/hanging/depth/disco.mp4" type="video/mp4"> 1147 Your browser does not support the video tag. 1148 </video> 1149 <img src="static/images/6x_black.svg" class="six_baseline" > 1150 <img src="static/images/cross.svg" class="tick-mark" > 1151 </div> 1152 1153 </div> 1154 <div class="video-container"> 1155 <div class="text-baseline"> 1156 <p>Point Cloud</p> 1157 </div> 1158 <div class="vid-tex-overlay"> 1159 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1160 <source src="static/videos/baselines/hanging/pointcloud/disco.mp4" type="video/mp4"> 1161 Your browser does not support the video tag. 1162 </video> 1163 <img src="static/images/6x_black.svg" class="six_baseline" > 1164 <img src="static/images/cross.svg" class="tick-mark" >
1165 </div> 1166 </div> 1167 </div> 1168 </div> 1169 </section> 1170 </div> 1171 </div> 1172 </div> 1173 </div> 1174 </div> 1175 </section> 1176 </div> 1177 1178 <!--------------------------------------------- Grape-Oven -------------------------------------------------------> 1179 1180 1181 <div class="container video-gallery-wrapper is-hidden" id="grape-oven"> 1182 1183 <!-- Video Gallery Section --> 1184 <section class="section is-light video-gallery" > 1185 1186 <!-- Task Description --> 1187 <!-- <h4 class="is-5 task-title";> 1188 <strong>Towel Folding</strong> 1189 </h4> --> 1190 1191 <!-- Combined Video Demonstration --> 1192 <div id="task-teaser"> 1193 <!-- <div class="video-container"> --> 1194 <!-- <video loading="lazy" preload="none" muted loop playsinline > 1195 <source src="static/videos/kp_filtering/towel_combined.mp4" type="video/mp4"> 1196 Your browser does not support the video tag. 1197 </video> --> 1198 <!-- </div> --> 1199 <div class="teaser-row"> 1200 <!-- Left Video --> 1201 <div class="video-container"> 1202 <h4 class="text-teaser";> 1203 Automatic Task-driven Keypoints Filteringâ </h4> 1204 <video loading="lazy" preload="none" muted loop playsinline > 1205 <source src="static/videos/teaser/oven_filtering.mp4" type="video/mp4"> 1206 Your browser does not support the video tag. 1207 </video> 1208 </div> 1209 1210 <!-- Center Image --> 1211 <div class="video-container"> 1212 <h4 class="text-teaser";> 1213 Keypoints Transfer during Inferenceââ </h4> 1214 <div class="image-grid"> 1215 <div class="image-container"> 1216 <img src="static/images/oven_2.png" alt="Image 1" class="teaser-image"> 1217 </div> 1218 <div class="image-container"> 1219 <img src="static/images/oven_3.png" alt="Image 2" class="teaser-image"> 1220 </div> 1221 <div class="image-container"> 1222 <img src="static/images/oven_4.png" alt="Image 3" class="teaser-image"> 1223 </div> 1224 <div class="image-container"> 1225 <img src="static/images/oven_5.png" alt="Image 4" class="teaser-image"> 1226 </div> 1227 </div> 1228 </div> 1229 1230 <!-- Right Video --> 1231 <div> 1232 <h4 class="text-teaser";> 1233 Keypoints Tracking during Rolloutâ </h4> 1234 <div class="video-container"> 1235 <div class="vid-tex-overlay"> 1236 <video loading="lazy" preload="none" muted loop playsinline > 1237 <source src="static/videos/teaser/oven_tracking.mp4" type="video/mp4"> 1238 Your browser does not support the video tag. 1239 </video> 1240 <img src="static/images/8x_black.svg" class="eight-teaser" > 1241 </div> 1242 </div> 1243 </div> 1244 </div> 1245 </div> 1246 1247 1248 <div class="container four-condn-container"> 1249 1250 <!-- Video Content Container --> 1251 <div class="video-content"> 1252 <!-- Side Views (DJI & Moving) --> 1253 <div class="view-content" id="side-views"> 1254 <div class="side-views-container"> 1255 <!-- DJI Camera View | Moving View--> 1256 <div id="dji-view-grape-oven"> 1257 <!-- Task 1 --> 1258 <div class="task-section"> 1259 <!-- Row 1 --> 1260 1261 1262 <!-- Task 1: Towel Folding --> 1263 <section class="section task-section"> 1264 <div class="condition-section"> 1265 <div class="video-selector-wrapper"> 1266 <div class="video-selector-container"> 1267 <div class="select mb-2"> 1268 <select class="video-selector" data-condition="random-position" data-view="dji"> 1269 <option value="1">Rollout 1</option> 1270 <option value="2">Rollout 2</option> 1271 <option value="3">Rollout 3</option> 1272 <option value="4">Rollout 4</option> 1273 1274 </select> 1275 1276 1277 </div> 1278 1279 </div> 1280 <div class="video-selector-condn"> 1281 <h5 class="subtitle is-5">Random Position</h5> 1282 </div> 1283 1284 </div> 1285 1286 <div class="videos-row"> 1287 <!-- First Row: Two Videos --> 1288 <div class="video-container"> 1289 <div class="text-atk"> 1290 <p>ATK (static view)</p> 1291 </div> 1292 <div class="vid-tex-overlay"> 1293 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1294 <source src="static/videos/dji/oven/rp1.mp4" type="video/mp4"> 1295 Your browser does not support the video tag. 1296 </video> 1297 <img src="static/images/6x_black.svg" class="six" > 1298 </div> 1299 1300 </div> 1301 <div class="video-container"> 1302 <div class="text-atk"> 1303 <p>ATK (moving view)</p> 1304 </div> 1305 <div class="vid-tex-overlay"> 1306 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1307 <source src="static/videos/iphone/oven/rp1.mp4" type="video/mp4"> 1308 Your browser does not support the video tag. 1309 </video> 1310 <img src="static/images/6x_black.svg" class="six" > 1311 </div> 1312 </div> 1313 </div> 1314 1315 1316 <div class="videos-row-baselines"> 1317 <!-- Second Row: Three Videos --> 1318 <div class="video-container success"> 1319 <div class="text-baseline"> 1320 <p>R3M</p> 1321 </div> 1322 <div class="vid-tex-overlay"> 1323 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1324 <source src="static/videos/baselines/oven/r3m/org.mp4" type="video/mp4"> 1325 Your browser does not support the video tag. 1326 </video> 1327 <img src="static/images/tick.svg" class="tick-mark" > 1328 <img src="static/images/6x_black.svg" class="six_baseline" > 1329 </div> 1330 1331 </div> 1332 1333 <div class="video-container"> 1334 <div class="text-baseline"> 1335 <p>Depth</p> 1336 </div> 1337 <div class="vid-tex-overlay"> 1338 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1339 <source src="static/videos/baselines/oven/depth/org.mp4" type="video/mp4"> 1340 Your browser does not support the video tag. 1341 </video> 1342 <img src="static/images/cross.svg" class="tick-mark" > 1343 <img src="static/images/6x_black.svg" class="six_baseline" > 1344 </div> 1345 1346 </div> 1347 <div class="video-container"> 1348 <div class="text-baseline"> 1349 <p>Point Cloud</p> 1350 </div> 1351 <div class="vid-tex-overlay"> 1352 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1353 <source src="static/videos/baselines/oven/pointcloud/org.mp4" type="video/mp4"> 1354 Your browser does not support the video tag. 1355 </video> 1356 <img src="static/images/cross.svg" class="tick-mark" > 1357 <img src="static/images/6x_black.svg" class="six_baseline" > 1358 </div> 1359 1360 </div> 1361 </div> 1362 </div> 1363 <div class="condition-section"> 1364 <div class="video-selector-wrapper"> 1365 <div class="video-selector-container"> 1366 <div class="select mb-2"> 1367 <select class="video-selector" data-condition="random-distractors" data-view="dji"> 1368 <option value="1">Rollout 1</option> 1369 <option value="2">Rollout 2</option> 1370 <option value="3">Rollout 3</option> 1371 <option value="4">Rollout 4</option> 1372 </select> 1373 </div> 1374 1375 </div> 1376 <div class="video-selector-condn"> 1377 <h5 class="subtitle is-5">Random Distractors</h5> 1378 </div> 1379 1380 </div> 1381 1382 <div class="videos-row"> 1383 <!-- First Row: Two Videos --> 1384 <div class="video-container"> 1385 <div class="text-atk"> 1386 <p>ATK (static view)</p> 1387 </div> 1388 <div class="vid-tex-overlay"> 1389 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1390 <source src="static/videos/dji/oven/rd1.mp4" type="video/mp4"> 1391 Your browser does not support the video tag. 1392 </video> 1393 <img src="static/images/6x_black.svg" class="six" > 1394 </div> 1395 1396 </div> 1397 <div class="video-container"> 1398 <div class="text-atk"> 1399 <p>ATK (moving view)</p> 1400 </div> 1401 <div class="vid-tex-overlay"> 1402 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1403 <source src="static/videos/iphone/oven/rd1.mp4" type="video/mp4"> 1404 Your browser does not support the video tag. 1405 </video> 1406 <img src="static/images/6x_black.svg" class="six" > 1407 </div> 1408 </div> 1409 </div> 1410 1411 1412 <div class="videos-row-baselines"> 1413 <!-- Second Row: Three Videos --> 1414 <div class="video-container success"> 1415 <div class="text-baseline"> 1416 <p>R3M</p> 1417 </div> 1418 <div class="vid-tex-overlay"> 1419 <video loading="lazy" p
1419reload="none" muted loop playsinline class="interactive-video"> 1420 <source src="static/videos/baselines/oven/r3m/dist.mp4" type="video/mp4"> 1421 Your browser does not support the video tag. 1422 </video> 1423 <img src="static/images/cross.svg" class="tick-mark" > 1424 <img src="static/images/6x_black.svg" class="six_baseline" > 1425 1426 </div> 1427 1428 </div> 1429 1430 <div class="video-container"> 1431 <div class="text-baseline"> 1432 <p>Depth</p> 1433 </div> 1434 <div class="vid-tex-overlay"> 1435 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1436 <source src="static/videos/baselines/oven/depth/dist.mp4" type="video/mp4"> 1437 Your browser does not support the video tag. 1438 </video> 1439 <img src="static/images/6x_black.svg" class="six_baseline" > 1440 <img src="static/images/cross.svg" class="tick-mark" > 1441 </div> 1442 1443 </div> 1444 <div class="video-container"> 1445 <div class="text-baseline"> 1446 <p>Point Cloud</p> 1447 </div> 1448 <div class="vid-tex-overlay"> 1449 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1450 <source src="static/videos/baselines/oven/pointcloud/dist.mp4" type="video/mp4"> 1451 Your browser does not support the video tag. 1452 </video> 1453 <img src="static/images/6x_black.svg" class="six_baseline" > 1454 <img src="static/images/cross.svg" class="tick-mark" > 1455 </div> 1456 1457 </div> 1458 </div> 1459 </div> 1460 <div class="condition-section"> 1461 <div class="video-selector-wrapper"> 1462 <div class="video-selector-container"> 1463 <div class="select mb-2"> 1464 <select class="video-selector" data-condition="random-backgrounds" data-view="dji"> 1465 <option value="1">Rollout 1</option> 1466 <option value="2">Rollout 2</option> 1467 <option value="3">Rollout 3</option> 1468 <option value="4">Rollout 4</option> 1469 </select> 1470 </div> 1471 1472 </div> 1473 <div class="video-selector-condn"> 1474 <h5 class="subtitle is-5">
1474Random Backgrounds</h5> 1475 </div> 1476 1477 </div> 1478 1479 <div class="videos-row"> 1480 <!-- First Row: Two Videos --> 1481 <div class="video-container"> 1482 <div class="text-atk"> 1483 <p>ATK (static view)</p> 1484 </div> 1485 <div class="vid-tex-overlay"> 1486 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1487 <source src="static/videos/dji/oven/rb1.mp4" type="video/mp4"> 1488 Your browser does not support the video tag. 1489 </video> 1490 <img src="static/images/6x_black.svg" class="six" > 1491 </div> 1492 1493 </div> 1494 <div class="video-container"> 1495 <div class="text-atk"> 1496 <p>ATK (moving view)</p> 1497 </div> 1498 <div class="vid-tex-overlay"> 1499 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1500 <source src="static/videos/iphone/oven/rb1.mp4" type="video/mp4"> 1501 Your browser does not support the video tag. 1502 </video> 1503 <img src="static/images/6x_black.svg" class="six" > 1504 </div> 1505 </div> 1506 </div> 1507 1508 1509 <div class="videos-row-baselines"> 1510 <!-- Second Row: Three Videos --> 1511 <div class="video-container success"> 1512 <div class="text-baseline"> 1513 <p>R3M</p> 1514 </div> 1515 <div class="vid-tex-overlay"> 1516 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1517 <source src="static/videos/baselines/oven/r3m/bck.mp4" type="video/mp4"> 1518 Your browser does not support the video tag. 1519 </video> 1520 <img src="static/images/6x_black.svg" class="six_baseline" > 1521 <img src="static/images/cross.svg" class="tick-mark" > 1522 </div> 1523 1524 </div> 1525 1526 <div class="video-container"> 1527 <div class="text-baseline"> 1528 <p>Depth</p> 1529 </div> 1530 <div class="vid-tex-overlay"> 1531 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1532 <source src="static/videos/baselines/oven/depth/bck.mp4" type="video/mp4"> 1533 Your browser does not support the video tag. 1534 </video> 1535 <img src="static/images/6x_black.svg" class="six_baseline" > 1536 <img src="static/images/cross.svg" class="tick-mark" > 1537 </div> 1538 1539 </div> 1540 <div class="video-container"> 1541 <div class="text-baseline"> 1542 <p>Point Cloud</p> 1543 </div> 1544 <div class="vid-tex-overlay"> 1545 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1546 <source src="static/videos/baselines/oven/pointcloud/bck.mp4" type="video/mp4"> 1547 Your browser does not support the video tag. 1548 </video> 1549 <img src="static/images/6x_black.svg" class="six_baseline" > 1550 <img src="static/images/cross.svg" class="tick-mark" > 1551 </div> 1552 1553 </div> 1554 </div> 1555 </div> 1556 1557 <div class="condition-section"> 1558 <div class="video-selector-wrapper"> 1559 <div class="video-selector-container"> 1560 <div class="select mb-2"> 1561 <select class="video-selector" data-condition="random-lights" data-view="dji"> 1562 <option value="1">Rollout 1</option> 1563 <option value="2">Rollout 2</option> 1564 </select> 1565 </div> 1566 1567 </div> 1568 <div class="video-selector-condn"> 1569 <h5 class="subtitle is-5">Random Lights</h5> 1570 </div> 1571 1572 </div> 1573 1574 <div class="videos-row"> 1575 <!-- First Row: Two Videos --> 1576 <div class="video-container"> 1577 <div class="text-atk"> 1578 <p>ATK (static view)</p> 1579 </div> 1580 <div class="vid-tex-overlay"> 1581 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1582 <source src="static/videos/dji/oven/disco.mp4" type="video/mp4"> 1583 Your browser does not support the video tag. 1584 </video> 1585 <img src="static/images/6x_black.svg" class="six" > 1586 </div> 1587 1588 </div> 1589 <div class="video-container"> 1590 <div class="text-atk"> 1591 <p>ATK (moving view)</p> 1592 </div> 1593 <div class="vid-tex-overlay"> 1594 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1595 <source src="static/videos/iphone/oven/disco.mp4" type="video/mp4"> 1596 Your browser does not support the video tag. 1597 </video> 1598 <img src="static/images/6x_black.svg" class="six" > 1599 </div> 1600 </div> 1601 </div> 1602 1603 1604 <div class="videos-row-baselines"> 1605 <!-- Second Row: Three Videos --> 1606 <div class="video-container success"> 1607 <div class="text-baseline"> 1608 <p>R3M</p> 1609 </div> 1610 <div class="vid-tex-overlay"> 1611 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1612 <source src="static/videos/baselines/oven/r3m/disco.mp4" type="video/mp4"> 1613 Your browser does not support the video tag. 1614 </video> 1615 <img src="static/images/6x_black.svg" class="six_baseline" > 1616 <img src="static/images/cross.svg" class="tick-mark" > 1617 </div> 1618 1619 </div> 1620 1621 <div class="video-container"> 1622 <div class="text-baseline"> 1623 <p>Depth</p> 1624 </div> 1625 <div class="vid-tex-overlay"> 1626 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1627 <source src="static/videos/baselines/oven/depth/disco.mp4" type="video/mp4"> 1628 Your browser does not support the video tag. 1629 </video> 1630 <img src="static/images/6x_black.svg" class="six_baseline" > 1631 <img src="static/images/cross.svg" class="tick-mark" > 1632 </div> 1633 1634 </div> 1635 <div class="video-container"> 1636 <div class="text-baseline"> 1637 <p>Point Cloud</p> 1638 </div> 1639 <div class="vid-tex-overlay"> 1640 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1641 <source src="static/videos/baselines/oven/pointcloud/disco.mp4" type="video/mp4"> 1642 Your browser does not support the video tag. 1643 </video> 1644 <img src="static/images/6x_black.svg" class="six_baseline" > 1645 <img src="static/images/cross.svg" class="tick-mark" > 1646 </div> 1647 </div> 1648 </div> 1649 </div> 1650 </section> 1651 </div> 1652 </div> 1653 </div> 1654 </div> 1655 </div> 1656 </section> 1657 </div> 1658<!--------------------------------------------- Pan Filling -------------------------------------------------------> 1659 1660 1661<div class="container video-gallery-wrapper is-hidden" id="pan-filling"> 1662 1663 <!-- Video Gallery Section --> 1664 <section class="section is-light video-gallery" > 1665 1666 <!-- Task Description --> 1667 <!-- <h4 class="is-5 task-title";> 1668 <strong>Towel Folding</strong> 1669 </h4> --> 1670 1671 <!-- Combined Video Demonstration --> 1672 <div id="task-teaser"> 1673 <!-- <div class="video-container"> --> 1674 <!-- <video loading="lazy" preload="none" muted loop playsinline > 1675 <source src="static/videos/kp_filtering/towel_c
1675ombined.mp4" type="video/mp4"> 1676 Your browser does not support the video tag. 1677 </video> --> 1678 <!-- </div> --> 1679 <div class="teaser-row"> 1680 <!-- Left Video --> 1681 <div class="video-container"> 1682 <h4 class="text-teaser";> 1683 Automatic Task-driven Keypoints Filteringâ </h4> 1684 <video loading="lazy" preload="none" muted loop playsinline > 1685 <source src="static/videos/teaser/pan_filtering.mp4" type="video/mp4"> 1686 Your browser does not support the video tag. 1687 </video> 1688 </div> 1689 1690 <!-- Center Image --> 1691 <div class="video-container"> 1692 <h4 class="text-teaser";> 1693 Keypoints Transfer during Inferenceââ </h4> 1694 <div class="image-grid"> 1695 <div class="image-container"> 1696 <img src="static/images/pan_2.png" alt="Image 1" class="teaser-image"> 1697 </div> 1698 <div class="image-container"> 1699 <img src="static/images/pan_3.png" alt="Image 2" class="teaser-image"> 1700 </div> 1701 <div class="image-container"> 1702 <img src="static/images/pan_4.png" alt="Image 3" class="teaser-image"> 1703 </div> 1704 <div class="image-container"> 1705 <img src="static/images/pan_5.png" alt="Image 4" class="teaser-image"> 1706 </div> 1707 </div> 1708 </div> 1709 1710 <!-- Right Video --> 1711 <div> 1712 <h4 class="text-teaser";> 1713 Keypoints Tracking during Rolloutâ </h4> 1714 <div class="video-container"> 1715 <div class="vid-tex-overlay"> 1716 <video loading="lazy" preload="none" muted loop playsinline > 1717 <source src="static/videos/teaser/pan_tracking.mp4" type="video/mp4"> 1718 Your browser does not support the video tag. 1719 </video> 1720 <img src="static/images/8x_black.svg" class="eight-teaser" > 1721 </div> 1722 </div> 1723 </div> 1724 </div> 1725 </div> 1726 1727 1728 <div class="container four-condn-container"> 1729 1730 <!-- Video Content Container --> 1731 <div class="video-content"> 1732 <!-- Side Views (DJI & Moving) --> 1733 <div class="view-content" id="side-views"> 1734 <div class="side-views-container"> 1735 <!-- DJI Camera View | Moving View--> 1736 <div id="dji-view-pan-filling"> 1737 <!-- Task 1 --> 1738 <div class="task-section"> 1739 <!-- Row 1 --> 1740 1741 1742 1743 <section class="section task-section"> 1744 <div class="condition-section"> 1745 <div class="video-selector-wrapper"> 1746 <div class="video-selector-container"> 1747 <div class="select mb-2"> 1748 <select class="video-selector" data-condition="random-position" data-view="dji"> 1749 <option value="1">Rollout 1</option> 1750 <option value="2">Rollout 2</option> 1751 <option value="3">Rollout 3</option> 1752 <option value="4">Rollout 4</option> 1753 1754 </select> 1755 1756 1757 </div> 1758 1759 </div> 1760 <div class="video-selector-condn"> 1761 <h5 class="subtitle is-5">Random Position</h5> 1762 </div> 1763 1764 </div> 1765 1766 <div class="videos-row"> 1767 <!-- First Row: Two Videos --> 1768 <div class="video-container"> 1769 <div class="text-atk"> 1770 <p>ATK (static view)</p> 1771 </div> 1772 <div class="vid-tex-overlay"> 1773 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1774 <source src="static/videos/dji/pan/rp1.mp4" type="video/mp4"> 1775 Your browser does not support the video tag. 1776 </video> 1777 <img src="static/images/6x_black.svg" class="six" > 1778 </div> 1779 1780 </div> 1781 <div class="video-container"> 1782 <div class="text-atk"> 1783 <p>ATK (moving view)</p> 1784 </div> 1785 <div class="vid-tex-overlay"> 1786 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1787 <source src="static/videos/iphone/pan/rp1.mp4" type="video/mp4"> 1788 Your browser does not support the video tag. 1789 </video> 1790 <img src="static/images/6x_black.svg" class="six" > 1791 </div> 1792 </div> 1793 </div> 1794 1795 1796 <div class="videos-row-baselines"> 1797 <!-- Second Row: Three Videos --> 1798 <div class="video-container success"> 1799 <div class="text-baseline"> 1800 <p>R3M</p> 1801 </div> 1802 <div class="vid-tex-overlay"> 1803 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1804 <source src="static/videos/baselines/pan/r3m/org.mp4" type="video/mp4"> 1805 Your browser does not support the video tag. 1806 </video> 1807 <img src="static/images/tick.svg" class="tick-mark" > 1808 <img src="static/images/6x_black.svg" class="six_baseline" > 1809 </div> 1810 1811 </div> 1812 1813 <div class="video-container"> 1814 <div class="text-baseline"> 1815 <p>Depth</p> 1816 </div> 1817 <div class="vid-tex-overlay"> 1818 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1819 <source src="static/videos/baselines/pan/depth/org.mp4" type="video/mp4"> 1820 Your browser does not support the video tag. 1821 </video> 1822 <img src="static/images/cross.svg" class="tick-mark" > 1823 <img src="static/images/6x_black.svg" class="six_baseline" > 1824 </div> 1825 1826 </div> 1827 <div class="video-container"> 1828 <div class="text-baseline"> 1829 <p>Point Cloud</p> 1830 </div> 1831 <div class="vid-tex-overlay"> 1832 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1833 <source src="static/videos/baselines/pan/pointcloud/org.mp4" type="video/mp4"> 1834 Your browser does not support the video tag. 1835 </video> 1836 <img src="static/images/cross.svg" class="tick-mark" > 1837 <img src="static/images/6x_black.svg" class="six_baseline" > 1838 </div> 1839 1840 </div> 1841 </div> 1842 </div> 1843 <div class="condition-section"> 1844 <div class="video-selector-wrapper"> 1845 <div class="video-selector-container"> 1846 <div class="select mb-2"> 1847 <select class="video-selector" data-condition="random-distractors" data-view="dji"> 1848 <option value="1">Rollout 1</option> 1849 <option value="2">Rollout 2</option> 1850 <option value="3">Rollout 3</option> 1851 <option value="4">Rollout 4</option> 1852 </select> 1853 </div> 1854 1855 </div> 1856 <div class="video-selector-condn"> 1857 <h5 class="subtitle is-5">Random Distractors</h5> 1858 </div> 1859 1860 </div> 1861 1862 <div class="videos-row"> 1863 <!-- First Row: Two Videos --> 1864 <div class="video-container"> 1865 <div class="text-atk"> 1866 <p>ATK (static view)</p> 1867 </div> 1868 <div class="vid-tex-overlay"> 1869 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1870 <source src="static/videos/dji/pan/rd1.mp4" type="video/mp4"> 1871 Your browser does not support the video tag. 1872 </video> 1873 <img src="static/images/6x_black.svg" class="six" > 1874 </div> 1875 1876 </div> 1877 <div class="video-container"> 1878 <div class="text-atk"> 1879 <p>ATK (moving view)</p> 1880 </div> 1881 <div class="vid-tex-overlay"> 1882 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1883 <source src="static/videos/iphone/pan/rd1.mp4" type="video/mp4"> 1884 Your browser does not support the video tag. 1885 </video> 1886 <img src="static/images/6x_black.svg" class="six" > 1887 </div> 1888 </div> 1889 </div> 1890 1891 1892 <div class="videos-row-baselines"> 1893 <!-- Second Row: Three Videos --> 1894 <div class="video-container success"> 1895 <div class="text-baseline"> 1896 <p>R3M</p> 1897 </div> 1898 <div class="vid-tex-overlay"> 1899 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1900 <source src="static/videos/baselines/pan/r3m/dist.mp4" type="video/mp4"> 1901 Your browser does not support the video tag. 1902 </video> 1903 <img src="static/images/cross.svg" class="tick-mark" > 1904 <img src="static/images/6x_black.svg" class="six_baseline" > 1905 1906 </div> 1907 1908 </div> 1909 1910 <div class="video-container"> 1911 <div class="text-baseline"> 1912 <p>Depth</p> 1913 </div> 1914 <div class="vid-tex-overlay"> 1915 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1916 <source src="static/videos/baselines/pan/depth/dist.mp4" type="video/mp4"> 1917 Your browser does not support the video tag. 1918 </video> 1919 <img src="static/images/6x_black.svg" class="six_baseline" > 1920 <img src="static/images/cross.svg" class="tick-mark" > 1921 </div> 1922 1923 </div> 1924 <div class="video-container"> 1925 <div class="text-baseline"> 1926 <p>Point Cloud</p> 1927
1927 </div> 1928 <div class="vid-tex-overlay"> 1929 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1930 <source src="static/videos/baselines/pan/pointcloud/dist.mp4" type="video/mp4"> 1931 Your browser does not support the video tag. 1932 </video> 1933 <img src="static/images/6x_black.svg" class="six_baseline" > 1934 <img src="static/images/cross.svg" class="tick-mark" > 1935 </div> 1936 1937 </div> 1938 </div> 1939 </div> 1940 <div class="condition-section"> 1941 <div class="video-selector-wrapper"> 1942 <div class="video-selector-container"> 1943 <div class="select mb-2"> 1944 <select class="video-selector" data-condition="random-backgrounds" data-view="dji"> 1945 <option value="1">Rollout 1</option> 1946 <option value="2">Rollout 2</option> 1947 <option value="3">Rollout 3</option> 1948 1949 </select> 1950 </div> 1951 1952 </div> 1953 <div class="video-selector-condn"> 1954 <h5 class="subtitle is-5">Random Backgrounds</h5> 1955 </div> 1956 1957 </div> 1958 1959 <div class="videos-row"> 1960 <!-- First Row: Two Videos --> 1961 <div class="video-container"> 1962 <div class="text-atk"> 1963 <p>ATK (static view)</p> 1964 </div> 1965 <div class="vid-tex-overlay"> 1966 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1967 <source src="static/videos/dji/pan/rb1.mp4" type="video/mp4"> 1968 Your browser does not support the video tag. 1969 </video> 1970 <img src="static/images/6x_black.svg" class="six" > 1971 </div> 1972 1973 </div> 1974 <div class="video-container"> 1975 <div class="text-atk"> 1976 <p>ATK (moving view)</p> 1977 </div> 1978 <div class="vid-tex-overlay"> 1979 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1980 <source src="static/videos/iphone/pan/rb1.mp4" type="video/mp4"> 1981 Your browser does not support the video tag. 1982 </video> 1983 <img src="static/images/6x_black.svg" class="six" > 1984 </div> 1985 </div> 1986 </div> 1987 1988 1989 <div class="videos-row-baselines"> 1990 <!-- Second Row: Three Videos --> 1991 <div class="video-container success"> 1992 <div class="text-baseline"> 1993 <p>R3M</p> 1994 </div> 1995 <div class="vid-tex-overlay"> 1996 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 1997 <source src="static/videos/baselines/pan/r3m/bck.mp4" type="video/mp4"> 1998 Your browser does not support the video tag. 1999 </video> 2000 <img src="static/images/6x_black.svg" class="six_baseline" > 2001 <img src="static/images/cross.svg" class="tick-mark" > 2002 </div> 2003 2004 </div> 2005 2006 <div class="video-container"> 2007 <div class="text-baseline"> 2008 <p>Depth</p> 2009 </div> 2010 <div class="vid-tex-overlay"> 2011 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2012 <source src="static/videos/baselines/pan/depth/bck.mp4" type="video/mp4"> 2013 Your browser does not support the video tag. 2014 </video> 2015 <img src="static/images/6x_black.svg" class="six_baseline" > 2016 <img src="static/images/cross.svg" class="tick-mark" > 2017 </div> 2018 2019 </div> 2020 <div class="video-container"> 2021 <div class="text-baseline"> 2022 <p>Point Cloud</p> 2023 </div> 2024 <div class="vid-tex-overlay"> 2025 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2026 <source src="static/videos/baselines/pan/pointcloud/bck.mp4" type="video/mp4"> 2027 Your browser does not support the video tag. 2028 </video> 2029 <img src="static/images/6x_black.svg" class="six_baseline" > 2030 <img src="static/images/cross.svg" class="tick-mark" > 2031 </div> 2032 2033 </div> 2034 </div> 2035 </div> 2036 2037 <div class="condition-section"> 2038 <div class="video-selector-wrapper"> 2039 <div class="video-selector-container"> 2040 <div class="select mb-2"> 2041 <select class="video-selector" data-condition="random-lights" data-view="dji"> 2042 <option value="1">Rollout 1</option> 2043 <option value="2">Rollout 2</option> 2044 </select> 2045 </div> 2046 2047 </div> 2048 <div class="video-selector-condn"> 2049 <h5 class="subtitle is-5">Random Lights</h5> 2050 </div> 2051 2052 </div> 2053 2054 <div class="videos-row"> 2055 <!-- First Row: Two Videos --> 2056 <div class="video-container"> 2057 <div class="text-atk"> 2058 <p>ATK (static view)</p> 2059 </div> 2060 <div class="vid-tex-overlay"> 2061 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2062 <source src="static/videos/dji/pan/disco.mp4" type="video/mp4"> 2063 Your browser does not support the video tag. 2064 </video> 2065 <img src="static/images/6x_black.svg" class="six" > 2066 </div> 2067 2068 </div> 2069 <div class="video-container"> 2070 <div class="text-atk"> 2071 <p>ATK (moving view)</p> 2072 </div> 2073 <div class="vid-tex-overlay"> 2074 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2075 <source src="static/videos/iphone/pan/disco.mp4" type="video/mp4"> 2076 Your browser does not support the video tag. 2077 </video> 2078 <img src="static/images/6x_black.svg" class="six" > 2079 </div> 2080 </div> 2081 </div> 2082 2083 2084 <div class="videos-row-baselines"> 2085 <!-- Second Row: Three Videos --> 2086 <div class="video-container success"> 2087 <div class="text-baseline"> 2088 <p>R3M</p> 2089 </div> 2090 <div class="vid-tex-overlay"> 2091 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2092 <source src="static/videos/baselines/pan/r3m/disco.mp4" type="video/mp4"> 2093 Your browser does not support the video tag. 2094 </video> 2095 <img src="static/images/6x_black.svg" class="six_baseline" > 2096 <img src="static/images/cross.svg" class="tick-mark" > 2097 </div> 2098 2099 </div> 2100 2101 <div class="video-container"> 2102 <div class="text-baseline"> 2103 <p>Depth</p> 2104 </div> 2105 <div class="vid-tex-overlay"> 2106 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2107 <source src="static/videos/baselines/pan/depth/disco.mp4" type="video/mp4"> 2108 Your browser does not support the video tag. 2109 </video> 2110 <img src="static/images/6x_black.svg" class="six_baseline" > 2111 <img src="static/images/cross.svg" class="tick-mark" > 2112 </div> 2113 2114 </div> 2115 <div class="video-container"> 2116 <div class="text-baseline"> 2117 <p>Point Cloud</p> 2118 </div> 2119 <div class="vid-tex-overlay"> 2120 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2121 <source src="static/videos/baselines/pan/pointcloud/disco.mp4" type="video/mp4"> 2122 Your browser does not support the video tag. 2123 </video> 2124 <img src="static/images/6x_black.svg" class="six_baseline" > 2125 <img src="static/images/cross.svg" class="tick-mark" > 2126 </div> 2127 </div> 2128 </div> 2129 </div> 2130 </section> 2131 </div> 2132 </div> 2133 </div> 2134 </div> 2135 </div> 2136 </section> 2137</div> 2138 2139 <!-- Results Section --> 2140<section class="section is-light result-container result-section"> 2141 <div class="container"> 2142 <h2 class="title has-text-centered" style="color: rgb(0, 0, 0);">Quantitative Results</h2> 2143 <div class="results"> 2144 <div class="result-columns is-centered"> 2145 <!-- Left Image --> 2146 <div class="column is-half"> 2147 <img src="static/images/Plot_imitation.png" alt="Plot Imitation" style="width: 100%; height: auto;"> 2148 </div> 2149 <!-- Right Image --> 2150 <div class="column is-half"> 2151 <img src="static/images/plot_imitation_atk.png" alt="Plot Imitation ATK" style="width: 100%; height: auto;"> 2152 </div> 2153 </div>
2154 <figcaption> 2155 <p><strong>Imitation policy success rates.</strong> <em>Left:</em> Aggregated results across diverse evaluation conditions show that <strong>ATK</strong> outperforms other methods based on different input modalities and selection strategies. <em>Right:</em> <strong>ATK</strong> demonstrates strong robustness under positional variation and various visual perturbations.</p> 2156 </figcaption> 2157 </div> 2158</section> 2159 2160 2161 <!-- Sim2Real --> 2162 <section class="section is-dark category-section "> 2163 <div class="container"> 2164 <h2 class="title has-text-centered" style="color: rgb(0, 0, 0);">2. Sim-to-real Transfer</h2> 2165 <p class="subtitle explain has-text-centered">ATK enables sim-to-real policies to remain robust 2166 and generalizable to environmental disturbance</p> 2167 2168 <div class="section is-light result-container result-section"> 2169 <div class="container"> 2170 <div class="results"> 2171 <div class="result is-centered"> 2172 <!-- Left Image --> 2173 <div class="sim2real-image"> 2174 <img src="static/images/sim2real.png" alt="Method" style="width: 100%; height: auto;"> 2175 </div> 2176 </div> 2177 <figcaption> 2178 2179 <p> <strong>Qualitative visualization of task-relevant keypoint selection and transfer.</strong> Keypoints 2180 selected in simulation transfer to real-world scenes across various object positions, backgrounds, 2181 distractors, and lighting.</p> 2182 </figcaption> 2183 </div> 2184 </div> 2185 </div> 2186 2187 <div class="container four-condn-container"> 2188 <!-- Sim2Real View --> 2189 <div class="view-content sim2real four"> 2190 2191 <div class="container"> 2192 2193 <!-- Video Content --> 2194 <div class="video-content"> 2195 <!-- Robot Camera View --> 2196 <div class="view-content"> 2197 <!-- Task 1 --> 2198 <div class="task-section"> 2199 2200 <!-- Row 1 --> 2201 <!-- Task 1: pan Folding --> 2202 <section class="section task-section"> 2203 <!-- <h5 class="title is-5 has-text-centered">Random Position</h5> --> 2204 <h4 class="is-5 has-text-centered task-title" style="display: flex;">Clock Hand Turning</h4> 2205 <div class="condition-section"> 2206 <div class="video-container"> 2207 <div class="videos-row"> 2208 <div class="column has-text-centered"> 2209 <!-- <h4 class="title is-5">Robot Camera </h4> --> 2210 <div class="vid-tex-overlay"> 2211 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2212 <source src="static/videos/sim2real/turn1.mp4" type="video/mp4"> 2213 Your browser does not support the video tag. 2214 </video> 2215 <img src="static/images/2x.svg" class="two" > 2216 </div> 2217 </div> 2218 <div class="column has-text-centered"> 2219 <!-- <h4 class="title is-5">DJI Camera </h4> --> 2220 <div class="vid-tex-overlay"> 2221 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2222 <source src="static/videos/sim2real/turn2.mp4" type="video/mp4"> 2223 Your browser does not support the video tag. 2224 </video> 2225 <img src="static/images/2x.svg" class="two" > 2226 </div> 2227 </div> 2228 <div class="column has-text-centered"> 2229 <!-- <h4 class="title is-5">Robot Camera</h4> --> 2230 <div class="vid-tex-overlay"> 2231 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2232 <source src="static/videos/sim2real/turn3.mp4" type="video/mp4"> 2233 Your browser does not support the video tag. 2234 </video> 2235 <img src="static/images/2x.svg" class="two" > 2236 </div> 2237 </div> 2238 </div> 2239 </div> 2240 </div> 2241 <!-- </section> 2242 <section class="section task-section"> --> 2243 <!-- <h5 class="title is-5 has-text-centered">Random Distractors</h5> --> 2244 <h5 class="is-5 has-text-centered task-title" style="display: flex;">Clock Button Press</h5> 2245 <div class="condition-section"> 2246 <div class="video-container"> 2247 <div class="videos-row"> 2248 <div class="column has-text-centered"> 2249 <!-- <h4 class="title is-5">Robot Camera </h4> --> 2250 <div class="vid-tex-overlay"> 2251 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2252 <source src="static/videos/sim2real/button1.mp4" type="video/mp4"> 2253 Your browser does not support the video tag. 2254 </video> 2255 <img src="static/images/2x.svg" class="two" > 2256 </div> 2257 </div> 2258 <div class="column has-text-centered"> 2259 <!-- <h4 class="title is-5">DJI Camera </h4> --> 2260 <div class="vid-tex-overlay"> 2261 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2262 <source src="static/videos/sim2real/button2.mp4" type="video/mp4"> 2263 Your browser does not support the video tag. 2264 </video> 2265 <img src="static/images/2x.svg" class="two" > 2266 </div> 2267 </div> 2268 <div class="column has-text-centered"> 2269 <!-- <h4 class="title is-5">Robot Camera</h4> --> 2270 <div class="vid-tex-overlay"> 2271 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2272 <source src="static/videos/sim2real/button3.mp4" type="video/mp4"> 2273 Your browser does not support the video tag. 2274 </video> 2275 <img src="static/images/2x.svg" class="two" > 2276 </div> 2277 </div> 2278 </div> 2279 </div> 2280 </div> 2281 <!-- </section> 2282 <section class="section task-section"> 2283 <h5 class="title is-5 has-text-centered">
2283Random Backgrounds</h5> --> 2284 <h5 class="is-5 has-text-centered task-title" style="display: flex;">GlassPot Lift</h5> 2285 <div class="condition-section"> 2286 <div class="video-container"> 2287 <div class="videos-row"> 2288 <div class="column has-text-centered"> 2289 <!-- <h4 class="title is-5">Robot Camera </h4> --> 2290 <div class="vid-tex-overlay"> 2291 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2292 <source src="static/videos/sim2real/glass1.mp4" type="video/mp4"> 2293 Your browser does not support the video tag. 2294 </video> 2295 <img src="static/images/2x.svg" class="two" > 2296 </div> 2297 </div> 2298 <div class="column has-text-centered"> 2299 <!-- <h4 class="title is-5">DJI Camera </h4> --> 2300 <div class="vid-tex-overlay"> 2301 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2302 <source src="static/videos/sim2real/glass2.mp4" type="video/mp4"> 2303 Your browser does not support the video tag. 2304 </video> 2305 <img src="static/images/2x.svg" class="two" > 2306 </div> 2307 </div> 2308 <div class="column has-text-centered"> 2309 <!-- <h4 class="title is-5">Robot Camera</h4> --> 2310 <div class="vid-tex-overlay"> 2311 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2312 <source src="static/videos/sim2real/glass3.mp4" type="video/mp4"> 2313 Your browser does not support the video tag. 2314 </video> 2315 <img src="static/images/2x.svg" class="two" > 2316 </div> 2317 </div> 2318 </div> 2319 </div> 2320 </div> 2321 <h5 class="is-5 has-text-centered task-title" style="display: flex;">Sushi Pick</h5> 2322 <div class="condition-section"> 2323 <div class="video-container"> 2324 <div class="videos-row"> 2325 <div class="column has-text-centered"> 2326 <!-- <h4 class="title is-5">Robot Camera </h4> --> 2327 <div class="vid-tex-overlay"> 2328 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2329 <source src="static/videos/sim2real/sushi1.mp4" type="video/mp4"> 2330 Your browser does not support the video tag. 2331 </video> 2332 <img src="static/images/2x.svg" class="two" > 2333 </div> 2334 </div> 2335 <div class="column has-text-centered"> 2336 <!-- <h4 class="title is-5">DJI Camera </h4> --> 2337 <div class="vid-tex-overlay"> 2338 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2339 <source src="static/videos/sim2real/sushi2.mp4" type="video/mp4"> 2340 Your browser does not support the video tag. 2341 </video> 2342 <img src="static/images/2x.svg" class="two" > 2343 </div> 2344 </div> 2345 <div class="column has-text-centered"> 2346 <!-- <h4 class="title is-5">Robot Camera</h4> --> 2347 <div class="vid-tex-overlay"> 2348 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2349 <source src="static/videos/sim2real/sushi3.mp4" type="video/mp4"> 2350 Your browser does not support the video tag. 2351 </video> 2352 <img src="static/images/2x.svg" class="two" > 2353 </div> 2354 </div> 2355 </div> 2356 </div> 2357 </div> 2358 </section> 2359 </div> 2360 </div> 2361 </div> 2362 </div> 2363 </div> 2364 2365 2366 2367 </div> 2368</div> 2369</section> 2370 2371 2372<!-- Results Section --> 2373<section class="section is-light result-container result-section"> 2374 2375 <h2 class="title has-text-centered" style="color: rgb(0, 0, 0);">Quantitative Results</h2> 2376 <div class="results"> 2377 <div class="result-columns is-centered"> 2378 <!-- Left Image --> 2379 <div class="column is-half"> 2380 <img src="static/images/plot_sim2real.png" alt="Plot Imitation" style="width: 100%; height: auto;"> 2381 </div> 2382 <!-- Right Image --> 2383 <div class="column is-half"> 2384 <img src="static/images/plot_sim2real_atk.png" alt="Plot Imitation ATK" style="width: 100%; height: auto;"> 2385 </div> 2386 </div>
2387 <figcaption> 2388 <p><strong>Sim-to-real policy success rates in real world.</strong> <em>Left:</em> Aggregated results across real-world evaluation conditionsârandom pose, background variation, distractor objects, and lighting changesâshow that <strong>ATK</strong> outperforms other methods using different input modalities. <em>Right:</em> <strong>ATK</strong> demonstrates strong robustness under positional variation and various visual perturbations.</p> 2389 </figcaption> 2390 </div> 2391 2392 2393 </div> 2394 2395</section> 2396<!-- Category Level Generalization --> 2397<section class="section is-dark category-section category-gen"> 2398 <div class="container"> 2399 <h2 class="title has-text-centered" style="color: rgb(0, 0, 0);">3. Category-Level Generalization</h2> 2400 <p class="subtitle explain has-text-centered" >Can our method generalize to different objects within the same category?</p> 2401 2402 <!-- <div class="content"> --> 2403 2404 2405 <section class="section task-section"> 2406 <!-- Task 1: Towel Folding --> 2407 <h4 class=" is-5 has-text-centered task-title"> 2408 Towel Folding 2409 </h4> 2410 <div class="videos-row"> 2411 <div class="column has-text-centered"> 2412 <h4 class="category-atk">Static View </h4> 2413 <div class="vid-tex-overlay"> 2414 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2415 <source src="static/videos/category-gen/towel/pink_towel_kp.mp4" type="video/mp4"> 2416 Your browser does not support the video tag. 2417 </video> 2418 <img src="static/images/6x_black.svg" class="six" > 2419 </div> 2420 </div> 2421 <div class="column has-text-centered"> 2422 <h4 class="category-atk">Moving View </h4> 2423 <div class="vid-tex-overlay"> 2424 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2425 <source src="static/videos/category-gen/towel/pink.mp4" type="video/mp4"> 2426 Your browser does not support the video tag. 2427 </video> 2428 <img src="static/images/6x_black.svg" class="six" > 2429 </div> 2430 </div> 2431 <div class="column has-text-centered"> 2432 <h4 class="category-atk">Static View</h4> 2433 <div class="vid-tex-overlay"> 2434 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2435 <source src="static/videos/category-gen/towel/blue_towel_kp.mp4" type="video/mp4"> 2436 Your browser does not support the video tag. 2437 </video> 2438 <img src="static/images/6x_black.svg" class="six" > 2439 </div> 2440 </div> 2441 <div class="column has-text-centered"> 2442 <h4 class="category-atk">Moving View</h4> 2443 <div class="vid-tex-overlay"> 2444 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2445 <source src="static/videos/category-gen/towel/blue.mp4" type="video/mp4"> 2446 Your browser does not support the video tag. 2447 </video> 2448 <img src="static/images/6x_black.svg" class="six" > 2449 </div> 2450 </div> 2451 </div> 2452 2453 2454 <!-- Task 2: Hanging Blanket --> 2455 <h4 class=" is-5 has-text-centered task-title"> 2456 Hanging Blanket 2457 </h4> 2458 2459 2460 <div class="videos-row"> 2461 <div class="column has-text-centered"> 2462 <h4 class="category-atk">Static View</h4> 2463 <div class="vid-tex-overlay"> 2464 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2465 <source src="static/videos/category-gen/hanging/pink_hanging_kp.mp4" type="video/mp4"> 2466 Your browser does not support the video tag. 2467 </video> 2468 <img src="static/images/6x_black.svg" class="six" > 2469 </div> 2470 </div> 2471 <div class="column has-text-centered"> 2472 <h4 class="category-atk">Moving View</h4> 2473 <div class="vid-tex-overlay"> 2474 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2475 <source src="static/videos/category-gen/hanging/pink.mp4" type="video/mp4"> 2476 Your browser does not support the video tag. 2477 </video> 2478 <img src="static/images/6x_black.svg" class="six" > 2479 </div> 2480 </div> 2481 <div class="column has-text-centered"> 2482 <h4 class="category-atk">Static View</h4> 2483 <div class="vid-tex-overlay"> 2484 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2485 <source src="static/videos/category-gen/hanging/blue_hanging_kp.mp4" type="video/mp4"> 2486 Your browser does not support the video tag. 2487 </video> 2488 <img src="static/images/6x_black.svg" class="six" > 2489 </div> 2490 </div> 2491 <div class="column has-text-centered"> 2492 <h4 class="category-atk">Moving View</h4> 2493 <div class="vid-tex-overlay"> 2494 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2495 <source src="static/videos/category-gen/hanging/blue.mp4" type="video/mp4"> 2496 Your browser does not support the video tag. 2497 </video> 2498 <img src="static/images/6x_black.svg" class="six" > 2499 </div> 2500 </div> 2501 </div> 2502 </section> 2503 2504 <!-- </div> --> 2505 </div> 2506 </section> 2507 <!-- End Category Level Generalization --> 2508 2509 2510 <!-- Viewpoint Generalization --> 2511<section class="section is-dark category-section viewpoint-gen"> 2512 <div class="container"> 2513 2514 <h2 class="title has-text-centered" style="color: rgb(0, 0, 0);">4. ViewPoint Generalization</h2> 2515 <p class="subtitle explain has-text-centered" >Can our method generalize to different camera viewpoints?</p> 2516 2517 <!-- <div class="content"> --> 2518 2519 <!-- Task 1: Towel Folding --> 2520 2521 <section class="section task-section"> 2522 <h4 class=" has-text-left task-title"> 2523 Towel Folding 2524 </h4> 2525 <div class="videos-row"> 2526 <div class="column has-text-centered"> 2527 <h4 class="text-atk">Viewpoint 1</h4> 2528 <div class="vid-tex-overlay"> 2529 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2530 <source src="static/videos/viewpoint-gen/towel/a.mp4" type="video/mp4"> 2531 Your browser does not support the video tag. 2532 </video> 2533 <img src="static/images/6x_black.svg" class="six" > 2534 </div> 2535 </div> 2536 <div class="column has-text-centered"> 2537 <h4 class="text-atk">Viewpoint 2</h4> 2538 <div class="vid-tex-overlay"> 2539 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2540 <source src="static/videos/viewpoint-gen/towel/b.mp4" type="video/mp4"> 2541 Your browser does not support the video tag. 2542 </video> 2543 <img src="static/images/6x_black.svg" class="six" > 2544 </div> 2545 </div> 2546 <div class="column has-text-centered"> 2547 <h4 class="text-atk">Viewpoint 3</h4> 2548 <div class="vid-tex-overlay"> 2549 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2550 <source src="static/videos/viewpoint-gen/towel/c.mp4" type="video/mp4"> 2551 Your browser does not support the video tag. 2552 </video> 2553 <img src="static/images/6x_black.svg" class="six" > 2554 </div> 2555 </div> 2556 </div> 2557 2558 2559 <!-- Task 2: Hanging Blanket --> 2560 <h4 class=" is-5 has-text-left task-title"> 2561 Hanging Blanket 2562 </h4> 2563 2564 <div class="videos-row"> 2565 <div class="column has-text-centered"> 2566 <h4 class="text-atk">Viewpoint 1</h4> 2567 <div class="vid-tex-overlay"> 2568 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2569 <source src="static/videos/viewpoint-gen/hanging/a.mp4" type="video/mp4"> 2570 Your browser does not support the video tag. 2571 </video> 2572 <img src="static/images/6x_black.svg" class="six" > 2573 </div> 2574 </div> 2575 <div class="column has-text-centered"> 2576 <h4 class="text-atk">Viewpoint 2</h4> 2577 <div class="vid-tex-overlay"> 2578 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2579 <source src="static/videos/viewpoint-gen/hanging/b.mp4" type="video/mp4"> 2580 Your browser does not support the video tag. 2581 </video> 2582 <img src="static/images/6x_black.svg" class="six" > 2583 </div> 2584 </div> 2585 <div class="column has-text-centered"> 2586 <h4 class="text-atk">Viewpoint 3</h4> 2587 <div class="vid-tex-overlay"> 2588 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2589 <source src="static/videos/viewpoint-gen/hanging/c.mp4" type="video/mp4"> 2590 Your browser does not support the video tag. 2591 </video> 2592 <img src="static/images/6x_black.svg" class="six" > 2593 </div> 2594 </div> 2595 </div> 2596 </section> 2597 2598 <!-- </div> --> 2599 </div> 2600 </section> 2601<!-- End of ViewPoint Generalization --> 2602 2603 2604<!-- Limitations Section --> 2605<section class="section is-light limitations-section"> 2606 <div class="container four-condn-container"> 2607 <h2 class="title has-text-centered " style="color: rgb(0, 0, 0);padding-bottom: 1%;">Failure Cases</h2> 2608 <p class="subtitle explain " > 2609 Using 2D keypoint pixel coordinates as input provides limited semantic information, which may limit the system's ability to fully understand the context of the task. Additionally, off-the-shelf tracking or correspondence modules may not always be robust enough for robotic applications, especially when dealing with uncommon or visually challenging data. In sim-to-real scenarios, discrepancies in dynamics between simulation and the real world further contribute to performance failures. 2610 </p> 2611 <h5 class=" is-5 has-text-left task-title" style="color: red;">Tracking Errors</h5> 2612 <div class="section task-section"> 2613 <div class="videos-row"> 2614 2615 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2616 <source src="static/videos/limitations/t_hanging.mp4" type="video/mp4"> 2617 Your browser does not support the video tag. 2618 </video> 2619 2620 2621 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2622 <source src="static/videos/limitations/t_pan.mp4" type="video/mp4"> 2623 Your browser does not support the video tag. 2624 </video> 2625 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2626 <source src="static/videos/limitations/t_shoe.mp4" type="video/mp4"> 2627 Your browser does not support the video tag. 2628 </video> 2629 2630 </div> 2631 </div> 2632 <h5 class=" is-5 has-text-left task-title" style="color: red;">Correspondense Mismatch</h5> 2633 <div class="section task-section"> 2634 <div class="videos-row"> 2635 <div class="video-container"> 2636 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2637 <source src="static/videos/limitations/c_hanging.mp4" type="video/mp4"> 2638 Your browser does not support the video tag. 2639 </video> 2640 </div> 2641 <div class="video-container"> 2642 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2643 <source src="static/videos/limitations/c2_hanging.mp4" type="video/mp4"> 2644 Your browser does not support the video tag. 2645 </video> 2646 </div> 2647 <div class="video-container"> 2648 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2649 <source src="static/videos/limitations/c_shoe.mp4" type="video/mp4"> 2650 Your browser does not support the video tag. 2651 </video> 2652 </div> 2653 </div> 2654 </div> 2655 2656 <h5 class=" is-5 has-text-left task-title" style="color: red;">Dynamics Gap</h5> 2657 2658 <div class="section task-section"> 2659 <div class="sim2real-videos-row"> 2660 <div class="videos-row"> 2661 <div class="video-container"> 2662 <h4 class="text-baseline">Simulation</h4> 2663 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2664 <source src="static/videos/limitations/sim_turn.mp4" type="video/mp4"> 2665 Your browser does not support the video tag. 2666 </video> 2667 </div> 2668 <div class="video-container"> 2669 <h4 class="text-baseline">Real World</h4> 2670 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2671 <source src="static/videos/limitations/real_turn.mp4" type="video/mp4"> 2672 Your browser does not support the video tag. 2673 </video> 2674 </div> 2675 </div> 2676 <div class="videos-row"> 2677 <div class="video-container"> 2678 <h4 class="text-baseline">Simulation</h4> 2679 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2680 <source src="static/videos/limitations/sim_button.mp4" type="video/mp4"> 2681 Your browser does not support the video tag. 2682 </video> 2683 </div> 2684 <div class="video-container"> 2685 <h4 class="text-baseline">Real World</h4> 2686 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2687 <source src="static/videos/limitations/real_button.mp4" type="video/mp4"> 2688 Your browser does not support the video tag. 2689 </video> 2690 </div> 2691 </div> 2692 </div> 2693 </div> 2694 <h5 class=" is-5 has-text-left task-title" style="color: red;">Representative Failure trajectories in the Real World</h5> 2695 2696 <div class="section task-section"> 2697 <div class="video-container limitations-combined-video"> 2698 <video loading="lazy" preload="none" muted loop playsinline class="interactive-video"> 2699
2699 <source src="static/videos/failure_main.mp4" type="video/mp4"> 2700 Your browser does not support the video tag. 2701 </video> 2702 </div> 2703 </div> 2704 </div> 2705 2706 2707 </section> 2708 <!-- End Limitations Section --> 2709 2710 2711 2712</div> 2713</div> 2714<!-- Team Section --> 2715<section class="section team-section"> 2716 <div class="container is-max-desktop"> 2717 <h2 class="title has-text-centered">Team</h2> 2718 <div class="team-grid"> 2719 <!-- Team Member 1 --> 2720 <div class="team-member"> 2721 <div class="image-container"> 2722 <img src="static/team/yunchu.jpg" alt="Yunchu Zhang" class="is-rounded"> 2723 </div> 2724 <p class="has-text-weight-bold mt-2 mb-0">Yunchu Zhang</p> 2725 <p class="is-size-7 ">University of Washington</p> 2726 </div> 2727 2728 <!-- Team Member 2 --> 2729 <div class="team-member"> 2730 <div class="image-container"> 2731 <img src="static/team/shubham.jpg" alt="Shubham Mittal" class="is-rounded"> 2732 </div> 2733 <p class="has-text-weight-bold mt-2 mb-0">Shubham Mittal</p> 2734 <p class="is-size-7">University of Washington</p> 2735 </div> 2736 2737 <!-- Team Member 3 --> 2738 <div class="team-member"> 2739 <div class="image-container"> 2740 <img src="static/team/octi.jpeg" alt="Zhengyu Zhang" class="is-rounded"> 2741 </div> 2742 <p class="has-text-weight-bold mt-2 mb-0">Zhengyu Zhang</p> 2743 <p class="is-size-7">University of Washington</p> 2744 </div> 2745 2746 <!-- Team Member 4 --> 2747 <div class="team-member"> 2748 <div class="image-container"> 2749 <img src="static/team/liyiming.png" alt="Liyiming Ke" class="is-rounded"> 2750 </div> 2751 <p class="has-text-weight-bold mt-2 mb-0">Liyiming Ke</p> 2752 <p class="is-size-7">University of Washington</p> 2753 </div> 2754 2755 <!-- Team Member 5 --> 2756 <div class="team-member"> 2757 <div class="image-container"> 2758 <img src="static/team/sidd.jpg" alt="Siddhartha Srinivasa" class="is-rounded"> 2759 </div> 2760 <p class="has-text-weight-bold mt-2 mb-0">Siddhartha Srinivasa</p> 2761 <p class="is-size-7">University of Washington</p> 2762 </div> 2763 2764 <!-- Team Member 6 --> 2765 <div class="team-member"> 2766 <div class="image-container"> 2767 <img src="static/team/abhishek.jpeg" alt="Abhishek Gupta" class="is-rounded"> 2768 </div> 2769 <p class="has-text-weight-bold mt-2 mb-0">Abhishek Gupta</p> 2770 <p class="is-size-7">University of Washington</p> 2771 </div> 2772 </div> 2773 </div> 2774</section> 2775<!-- End Team Section --> 2776 2777<section class="section" id="BibTeX" style="display: flex; justify-content: center; align-items: center;"> 2778 <div class="container is-max-desktop content has-text-centered"> 2779 <h2 class="title has-text-centered">BibTeX</h2> 2780 <pre style="text-align: center; overflow-x: auto; white-space: pre-wrap; word-wrap: break-word;"> 2781 <code style="text-align: left; display: inline-block; background-color: #fbfbfb; padding: 1rem; border-radius: 5px; box-shadow: 0 2px 5px rgba(0, 0, 0, 0.1);"> 2782@misc{zhang2025atkautomatictaskdrivenkeypoint, 2783 title={ATK: Automatic Task-driven Keypoint Selection for Robust Policy Learning}, 2784 author={Yunchu Zhang and Shubham Mittal and Zhengyu Zhang and Liyiming Ke and Siddhartha Srinivasa and Abhishek Gupta}, 2785 year={2025}, 2786 eprint={2506.13867}, 2787 archivePrefix={arXiv}, 2788 primaryClass={cs.RO}, 2789 url={https://arxiv.org/abs/2506.13867}, 2790} 2791 </code> 2792 </pre> 2793 </div> 2794</section> 2795 2796 <footer class="footer"> 2797 <div class="container"> 2798 <div class="columns is-centered"> 2799 <div class="column is-8"> 2800 <div class="content"> 2801 2802 <p> 2803 Website was built using the <a href="https://github.com/eliahuhorwitz/Academic-project-page-template" target="_blank">Academic Project Page Template</a> and the <a href="https://nerfies.github.io" target="_blank">Nerfies</a> project page. 2804 2805 </p> 2806 2807 </div> 2808 </div> 2809 </div> 2810 </div> 2811</footer> 2812</div> 2813 </body> 2814 </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.