-
Notifications
You must be signed in to change notification settings - Fork 1
/
Copy pathindex.html
735 lines (655 loc) · 44 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
<!DOCTYPE html>
<html>
<head>
<title>xTED</title>
<style>
.hidden {
display: none;
}
</style>
<script src="https://cdn.jsdelivr.net/npm/chart.js"></script>
<meta charset="utf-8">
<meta name="description"
content="Cross-Domain Adaptation via
Diffusion-Based Trajectory Editing">
<meta name="keywords" content="Cross-Embodiment Learning, Embodied AI, Diffusion Models">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Cross-Domain Adaptation via
Diffusion-Based Trajectory Editing</title>
<title>Cross-Domain Adaptation via
Diffusion-Based Trajectory Editing</title>
<!-- <link rel="icon" href="./assets/icon.png"> -->
<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro" rel="stylesheet">
<link rel="stylesheet" href="./assets/css/bulma.min.css">
<link rel="stylesheet" href="./assets/css/bulma-carousel.min.css">
<link rel="stylesheet" href="./assets/css/bulma-slider.min.css">
<link rel="stylesheet" href="./assets/css/fontawesome.all.min.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<script defer src="./assets/js/fontawesome.all.min.js"></script>
</head>
<body>
<section class="hero">
<div class="hero-body">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column has-text-centered">
<h1 class="title is-1 publication-title is-bold">
<!-- <img src="./assets/icon.png" style="width:1em;vertical-align: middle" alt="Logo"/> -->
<span class="mmmu" style="vertical-align: middle">xTED</span>
</h1>
<h2 class="subtitle is-3 publication-subtitle">
Cross-Domain Adaptation via Diffusion-Based Trajectory Editing
<!-- <br>
and Reasoning Benchmark for Expert AGI -->
</h2>
<div class="is-size-5 publication-authors">
<span class="author-block"><a href="https://t6-thu.github.io/" target="_blank">Haoyi Niu</a>* <sup>†</sup> <sup style="color:#4385F5;">1</sup>,</span>
<span class="author-block">Qimao Chen* <sup style="color:#4385F5;">1</sup>,</span>
<span class="author-block">Tenglong Liu <sup style="color:#F5B400;">2</sup> <sup style="color:#4385F5;">1</sup>,</span><br>
<span class="author-block">Jianxiong Li<sup style="color:#4385F5;">1</sup>,</span>
<span class="author-block">Guyue Zhou<sup style="color:#4385F5;">1</sup>,</span>
<span class="author-block">Yi Zhang<sup>✉</sup> <sup style="color:#4385F5;">1</sup>,</span>
<span class="author-block">Jianming Hu<sup>✉</sup> <sup style="color:#4385F5;">1</sup>,</span>
<!-- <span class="author-block">XXX<sup style="color:#F5B400;">4</sup> <sup style="color:#4385F5;">2</sup>,</span><br>
<span class="author-block">XXX<sup style="color:#4385F5;">2</sup>,</span>
<span class="author-block">XXX<sup style="color:#DC4437;">1</sup>,</span>
<span class="author-block">XXX<sup style="color:#DC4437;">1</sup>,</span> -->
<span class="author-block">Xianyuan Zhan<sup>✉</sup> <sup style="color:#4385F5;">1</sup> <sup style="color:#109D59;">3</sup>,</span>
</div>
<br>
<div class="is-size-5 publication-authors">
<span class="author-block"><sup style="color:#4385F5;">1</sup>Tsinghua University</span>
<span class="author-block"><sup style="color:#F5B400;">2</sup>National University of Defense Technology</span>
<span class="author-block"><sup style="color:#109D59;">3</sup>Shanghai AI Lab</span><br>
<!-- <span class="author-block"><sup style="color:#109D59;">3</sup>XXX</span><br>
<span class="author-block"><sup style="color:#F5B400;">4</sup>XXX</span>
<span class="author-block"><sup style="color:#f542dd;">5</sup>XXX</span> -->
</div>
<br>
<div class="is-size-5 publication-authors">
<span class="author-block">*Equal contribution, </span><br>
<span class="author-block">†Project Lead:</span>
<span class="author-block"><a href="mailto:nhy22@mails.tsinghua.edu.cn">nhy22@mails.tsinghua.edu.cn</a></span><br>
<span class="author-block">✉Correspondence to:</span>
<span class="author-block"><a href="mailto:zhanxianyuan@air.tsinghua.edu.cn">zhanxianyuan@air.tsinghua.edu.cn</a></span>
</div>
<!-- <style>
.accepted {
text-align: center;
color: #0a283d; /* Dark blue color */
font-size: 24px;
background-color: #e8f0fe79; /* Light blue background */
border: 1px solid #B6D4FE;
border-radius: 10px;
padding: 20px;
box-shadow: 0 4px 8px rgba(0,0,0,0.1);
}
</style>
<div class="accepted">
<i class="fas fa-fire icon" style="color: red;"></i>
Exciting News! Our paper has been accepted at ICML2024 and selected as outstanding paper at MFM-EAI workshop@ICML2024!
</div> -->
<div class="column has-text-centered">
<div class="publication-links">
<!-- Arxiv PDF Link. -->
<span class="link-block">
<a href="https://arxiv.org/abs/2409.08687"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-file-pdf"></i>
</span>
<span>Arxiv</span>
</a>
</span>
<!-- <span class="link-block">
<a href="https://openreview.net/forum?id=j5JvZCaDM0"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-file-pdf"></i>
</span>
<span>Openreview</span>
</a>
</span> -->
<!-- <span class="link-block">
<a href="https://cloud.tsinghua.edu.cn/d/0d2939f7f7234cf68f1d/"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<p style="font-size:18px">🔗</p>
</span>
<span>Dataset</span>
</a>
</span> -->
<span class="link-block">
<a href="https://github.com/t6-thu/xTED"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fab fa-github"></i>
</span>
<span>Code</span>
</a>
</span>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<style>
.center {
display: block;
margin-left: auto;
margin-right: auto;
width: 80%;
}
</style>
<!-- <section class="hero teaser">
<div class="container is-max-desktop">
<div class="hero-body">
<video id="matting-video" autoplay muted controls playsinline width="100%" >
<source src="./assets/web/teaser.mp4" type="video/mp4">
</video>
</div>
</div>
</section> -->
<section class="section">
<div class="container" style="margin-bottom: 2vh;">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h1 class="title is-2">Abstract</h1>
<div class="content has-text-justified">
<p>
Reusing pre-collected data from different domains is an appealing solution for decision-making tasks that have insufficient data in the target domain but are relatively abundant in other related domains. Existing cross-domain policy transfer methods mostly aim at learning domain correspondences or corrections to facilitate policy learning, such as learning domain/task-specific discriminators, representations, or policies. This design philosophy often results in heavy model architectures or task/domain-specific modeling, lacking flexibility. This reality makes us wonder: can we directly bridge the domain gaps universally at the data level, instead of relying on complex downstream cross-domain policy transfer models? In this study, we propose the <b>Cross</b>-Domain <b>T</b>rajectory <b>ED</b>iting (<b>xTED</b>) framework that employs a specially designed diffusion model for cross-domain trajectory adaptation. Our proposed model architecture effectively captures the intricate dependencies among states, actions, and rewards, as well as the dynamics patterns within target data. By utilizing the pre-trained diffusion as a prior, source domain trajectories can be transformed to match with target domain properties while preserving original semantic information. This process implicitly corrects underlying domain gaps, enhancing state realism and dynamics reliability in the source data, and allowing flexible incorporation with various downstream policy learning methods. Despite its simplicity, xTED demonstrates superior performance in extensive simulation and real-robot experiments.
</p>
</div>
</div>
</div>
</div>
</section>
<section class="section">
<div class="container">
<div class="columns is-centered has-text-centered">
<!-- <div class="column is-full-width has-text-centered"> -->
<div class="column is-four-fifths">
<h1 class="title is-2">Downstream Robot Manipulation Tasks</h1>
<div class="content has-text-justified">
<p>
We conduct real-world experiments in robotic environments where target data is collected by (a) WidowX robot and source data is collected by (b) Airbot, for 100 trajectories respectively. We build three manipulation tasks: (1) Picking up a red cup on a silver pan (Cup); (2) Picking up a duck on a green plate (Duck); (3) Moving a pot from right to left (Pot).
</p>
<!-- <div style="display: grid; grid-template-columns: repeat(5, 1fr); gap: 10px;">
<div style="text-align: center;">
<img src="./assets/web/robot/air.png" alt="WidowX" style="max-width: 100%;">
<p><i>(a) WidowX</i></p>
</div>
<div style="text-align: center;">
<img src="./assets/web/robot/air_pickredcup.jpg" alt="Primary View" style="max-width: 100%;">
<p><i>(b) Primary View</i></p>
</div>
<div style="text-align: center;">
<img src="./assets/web/robot/air_pickredcup1.jpg" alt="Gripper View" style="max-width: 100%;">
<p><i>(c) Gripper View</i></p>
</div>
<div style="text-align: center;">
<img src="./assets/web/robot/air_duck.jpg" alt="Primary View" style="max-width: 100%;">
<p><i>(d) Primary View</i></p>
</div>
<div style="text-align: center;">
<img src="./assets/web/robot/air_duck1.jpg" alt="Gripper View" style="max-width: 100%;">
<p><i>(e) Gripper View</i></p>
</div>
<div style="text-align: center;">
<img src="./assets/web/robot/airbot.png" alt="Airbot" style="max-width: 100%;">
<p><i>(f) Airbot</i></p>
</div>
<div style="text-align: center;">
<img src="./assets/web/robot/pick_redcup.jpg" alt="Primary View" style="max-width: 100%;">
<p><i>(g) Primary View</i></p>
</div>
<div style="text-align: center;">
<img src="./assets/web/robot/pick_redcup1.jpg" alt="Gripper View" style="max-width: 100%;">
<p><i>(h) Gripper View</i></p>
</div>
<div style="text-align: center;">
<img src="./assets/web/robot/pick_duck.jpg" alt="Primary View" style="max-width: 100%;">
<p><i>(i) Primary View</i></p>
</div>
<div style="text-align: center;">
<img src="./assets/web/robot/pick_duck_1.jpg" alt="Gripper View" style="max-width: 100%;">
<p><i>(j) Gripper View</i></p>
</div>
</div> -->
<div style="text-align: center;">
<div style="margin-bottom: 20px;">
<img src="./assets/web/xted-robot.png" alt="real world experiment" style="width: 100%; height: auto;">
</div>
</div>
<p style="text-align:center;">
<b>Figure 1:</b> The comparisons between the target domain and the source domain.
Target and source domains with complicated discrepancies on embodiments and viewpoints (top) and experiment results (bottom). The top right presents the snapshots from base and wrist camera views of data collection processes in target/source domain from <b>Cup</b>/<b>Duck</b>/<b>Pot</b> tasks respectively. The average success rate for real-robot tasks with/without distractors is obtained over 3 seeds.
</p>
<!-- <p style="text-align:center;">
<b>Figure 1:</b> The comparisons between the target domain and the source domain.
Figure (b-e) and (g-j) show the process of data collection in target and source domain respectively. For example, Figure (b,c,g,h) and (d,e,i,j) show the primary views and gripper views from <b>Cup</b> and <b>Duck</b> tasks respectively.
</p> -->
<div class="column">
<h2 style="text-align:left" class="title is-3">Results on Real Robots</h2>
<div class="content has-text-justified">
<!-- <div class="content has-text-centered">
<img src="./assets/web/redcup.png" alt="algebraic reasoning" class="center">
<p><b><i>Figure 1:</i></b> robot exprimental results. Success rate is averaged over 10 episodes and 3 seeds.</p>
</div> -->
<!-- <div style="display: flex; justify-content: center; align-items: center;">
<div style="margin: 0 5px;">
<img src="./assets/web/redcup.png" alt="Red cup on silver pan" style="max-width: 100%;">
</div>
<div style="margin: 0 5px;">
<img src="./assets/web/duck.png" alt="Duck on green plate" style="max-width: 100%;">
</div>
<div style="margin: 0 5px;">
<img src="./assets/web/move.png" alt="Move pot" style="max-width: 100%;">
</div>
</div>
<p style="text-align: center;"><b><i>Figure:</i></b> Real robot experimental results. Success rate is averaged over 10 episodes and 3 seeds.</p> -->
<div style="text-align: center;">
<div style="margin-bottom: 20px;">
<img src="./assets/web/redcup.png" alt="Red cup on silver pan" style="width: 100%; max-width: 600px;">
</div>
<div style="margin-bottom: 20px;">
<img src="./assets/web/duck.png" alt="Duck on green plate" style="width: 100%; max-width: 600px;">
</div>
<div style="margin-bottom: 20px;">
<img src="./assets/web/move.png" alt="Move pot" style="width: 100%; max-width: 600px;">
</div>
<p style="text-align: center;"><b><i>Figure:</i></b> Real robot experimental results. Success rate is averaged over 10 episodes and 3 seeds.</p>
</div>
<div class="column">
<h2 style="text-align:left" class="title is-3">Target + Edited Source (xTED)</h2>
<div class="content has-text-justified">
<div class="column">
<h3 style="text-align:left" class="title is-4">Red cup on silver pan</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/redcupsilverpan/redcupsilverpan1.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/redcupsilverpan/redcupsilverpan2.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/redcupsilverpan/redcupsilverpan3.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Red cup on silver pan with distraction</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/redcupsilverpan/redcupsilverpan9-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/redcupsilverpan/redcupsilverpan6-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/redcupsilverpan/redcupsilverpan3-d.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Duck on green plate</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/duckgreenplate/duckgreenplate4.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/duckgreenplate/duckgreenplate5.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/duckgreenplate/duckgreenplate6.mp4"
type="video/mp4">
</video>
</div>
<h3 style="text-align:left" class="title is-4">Duck on green plate with distraction</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/duckgreenplate/duckgreenplate7-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/duckgreenplate/duckgreenplate9-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/duckgreenplate/duckgreenplate10-d.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Move pot</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/movepot/movepot4.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/movepot/movepot6.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/movepot/movepot7.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Move pot with distraction</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/movepot/movepot1-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/movepot/movepot5-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/movepot/movepot4-d.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h2 style="text-align:left" class="title is-3">Target</h2>
<div class="content has-text-justified">
<div class="column">
<h3 style="text-align:left" class="title is-4">Red cup on silver pan</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/redcup/target_cup1.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/redcup/target_cup2.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/redcup/target_cup3.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Red cup on silver pan with distraction</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/redcup/d_target_cup3.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/redcup/d_target_cup5.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/redcup/d_target_cup4.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Duck on green plate</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/duck/target_duck1.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/duck/target_duck2.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/duck/target_duck3.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Duck on green plate with distraction</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/duck/d_target_duck1.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/duck/d_target_duck2.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/duck/d_target_duck3.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Move pot</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/move/target_move1.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/move/target_move2.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/move/target_move3.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Move pot with distraction</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/move/d_target_move1.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/move/d_target_move2.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target/move/d_target_move3.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h2 style="text-align:left" class="title is-3">Target + Source</h2>
<div class="content has-text-justified">
<div class="column">
<h3 style="text-align:left" class="title is-4">Red cup on silver pan</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/cup1.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/cup2.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/cup3.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Red cup on silver pan with distraction</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/cup1-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/cup2-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/cup3-d.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Duck on green plate</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/duck1.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/duck2.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/duck3.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Duck on green plate with distraction</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/duck1-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/duck2-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/duck3-d.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Move pot</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/move1.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/move2.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/move3.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<h3 style="text-align:left" class="title is-4">Move pot with distraction</h3>
<div class="content has-text-justified">
</div>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/move1-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/move2-d.mp4"
type="video/mp4">
</video>
<video id="dollyzoom" autoplay controls muted loop playsinline width="32.5%" height="100%">
<source src="./assets/web/target_unedit/move3-d.mp4"
type="video/mp4">
</video>
</div>
<!-- <div class="column">
<h2 style="text-align:left" class="title is-3">Results on Simulation</h2>
<div class="content has-text-justified">
<p>
We also evaluate on the FrankaKitchen benchmark.
We train LCBC policies on 5 tasks in FrankaKitchen environment using 1/3/5 demonstrations for each task.
DecisionNCE achieves the highest success rate across diverse dataset quantities,
demonstrating its effectiveness in extracting valuable information from out-of-domain data.
</p>
<div class="content has-text-centered">
<img src="./assets/web/simulation.png" alt="algebraic reasoning" class="center">
<p><b><i>Figure 2:</i></b> Simulation LCBC results. Max success rate averaged over 25 evaluation episodes and 3 seeds.</p>
</div>
<img src="./assets/web/simulation/1.gif" alt="Image" style="width: 19%;">
<img src="./assets/web/simulation/2.gif" alt="Image" style="width: 19%;">
<img src="./assets/web/simulation/3.gif" alt="Image" style="width: 19%;">
<img src="./assets/web/simulation/4.gif" alt="Image" style="width: 19%;">
<img src="./assets/web/simulation/5.gif" alt="Image" style="width: 19%;">
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</section> -->
<!-- <section class="section">
<div class="container">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Feasibility-Guided Diffusion Model</h2>
<div class="content has-text-justified">
<p>
We propose a feasibility-dependent objective, i.e.,
<b>maximizing reward value within the feasible region while minimizing safety risks in the infeasible region</b>.
In FISOR, the optimal policy for the optimization problem can be derived in a special form of weighted behavior cloning.
Moreover, we propose a novel energy-guided sampling method that <b>does not require training a complicated time-dependent classifier</b> to simplify the training. <b>No more Lagrangian.</b>
</p>
<div class="content has-text-centered">
<img src="./assets/framework.jpg" alt="algebraic reasoning" class="center">
<p><b><i>Figure 2:</i></b> Feasibility-guided diffusion model with time-independent classifier-guided sampling method.</p>
</div>
</div>
</div>
</div>
</section> -->
<!-- @PAN TODO: bibtex -->
<section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title is-3 has-text-centered">BibTeX</h2>
<pre><code>
@inproceedings{niu2024xted,
title={xTED: Cross-Domain Adaptation via Diffusion-Based Trajectory Editing},
author={Niu, Haoyi and Chen, Qimao and Liu, Tenglong and Li, Jianxiong and Zhou, Guyue and Zhang, Yi and Hu, Jianming and Zhan, Xianyuan},
booktitle={NeurIPS 2024 Workshop on Open-World Agents}
}
</code></pre>
</div>
</section>
<footer class="footer">
<!-- <div class="container"> -->
<div class="content has-text-centered">
</div>
<div class="columns is-centered">
<div class="column is-8">
<div class="content has-text-centered">
<p>
This website is website adapted from <a href="https://mmmu-benchmark.github.io/">MMMU</a>, licensed under a <a rel="license"
href="http://creativecommons.org/licenses/by-sa/4.0/">Creative
Commons Attribution-ShareAlike 4.0 International License</a>.
</p>
</div>
</div>
</div>
<!-- </div> -->
</footer>
</body>