-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
505 lines (462 loc) · 22.2 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
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8">
<!-- Meta tags for social media banners, these should be filled in appropriatly as they are your "business card" -->
<!-- Replace the content tag with appropriate information -->
<meta name="description" content="by Shivin Dass, Jiaheng Hu, Ben Abbatematteo, Peter Stone, Roberto Martín-Martín">
<meta property="og:title" content="Learning to Look: Seeking information for Decision Making via Policy Factorization"/>
<meta property="og:description" content="Learning to Look: Seeking information for Decision Making via Policy Factorization"/>
<meta property="og:url" content="https://robin-lab.cs.utexas.edu/learning2look/"/>
<!-- Path to banner image, should be in the path listed below. Optimal dimenssions are 1200X630-->
<!-- <meta property="og:image" content="static/images/telemoma_architecture.png" />
<meta property="og:image:width" content="1200"/>
<meta property="og:image:height" content="630"/> -->
<meta name="twitter:title" content="Learning to Look: Seeking information for Decision Making via Policy Factorization">
<meta name="twitter:description" content="Learning to Look: Seeking information for Decision Making via Policy Factorization">
<!-- Path to banner image, should be in the path listed below. Optimal dimenssions are 1200X600-->
<!-- <meta name="twitter:image" content="https://robin-lab.cs.utexas.edu/telemoma-web/static/images/telemoma_architecture.png">
<meta name="twitter:card" content="summary_large_image"> -->
<!-- Keywords for your paper to be indexed by-->
<meta name="keywords" content="Robotics, Active Vision, Interactive Perception">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Learning to Look</title>
<link rel="icon" type="image/x-icon" href="static/images/favicon.ico">
<link href="https://fonts.googleapis.com/css?family=Google+Sans|Noto+Sans|Castoro"
rel="stylesheet">
<link rel="stylesheet" href="static/css/bulma.min.css">
<link rel="stylesheet" href="static/css/bulma-carousel.min.css">
<link rel="stylesheet" href="static/css/bulma-slider.min.css">
<link rel="stylesheet" href="static/css/fontawesome.all.min.css">
<link rel="stylesheet"
href="https://cdn.jsdelivr.net/gh/jpswalsh/academicons@1/css/academicons.min.css">
<link rel="stylesheet" href="static/css/index.css">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script src="https://documentcloud.adobe.com/view-sdk/main.js"></script>
<script defer src="static/js/fontawesome.all.min.js"></script>
<script src="static/js/bulma-carousel.min.js"></script>
<script src="static/js/bulma-slider.min.js"></script>
<script src="static/js/index.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"><span style="color:#A52B16; font-weight: bold;">Learning to Look</span> 👀</h1>
<h2 class="subtitle is-2 publication-subtitle">Seeking information for Decision Making via Policy Factorization</h1>
<div class="is-size-5 publication-authors">
<!-- Paper authors -->
<span class="author-block">
<a href="https://shivindass.github.io/" target="_blank">Shivin Dass</a><sup>1</sup>,</span>
<a href="https://jiahenghu.github.io/" target="_blank">Jiaheng Hu</a><sup>1</sup>,</span>
<span class="author-block">
<a href="https://babbatem.github.io/" target="_blank">Ben Abbatematteo</a><sup>1</sup>,</span>
<span class="author-block">
<a href="https://www.cs.utexas.edu/~pstone/" target="_blank">Peter Stone</a><sup>1,2</sup>,</span>
<span class="author-block">
<a href="https://robertomartinmartin.com/" target="_blank">Roberto Martín-Martín</a><sup>1</sup></span>
</div>
<div class="is-size-5 publication-authors">
<span class="author-block"><sup>1</sup>The University of Texas at Austin, <sup>2</sup>Sony AI
<!-- <br>Conferance name and year -->
</span>
<!-- <span class="eql-cntrb"><small><br><sup>*</sup>Indicates Equal Contribution</small></span> -->
</div>
<div class="column has-text-centered">
<div class="publication-links">
<!-- Arxiv PDF link -->
<span class="link-block">
<a href="http://arxiv.org/abs/2410.18964" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fas fa-file-pdf"></i>
</span>
<span>Paper</span>
</a>
</span>
<!-- Github link -->
<span class="link-block">
<a href="" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="fab fa-github"></i>
</span>
<span>Code (Coming Soon!)</span>
</a>
</span>
<!-- ArXiv abstract Link -->
<!-- <span class="link-block">
<a href="" target="_blank"
class="external-link button is-normal is-rounded is-dark">
<span class="icon">
<i class="ai ai-arxiv"></i>
</span>
<span>arXiv</span>
</a> -->
<!-- </span> -->
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<!-- Teaser video-->
<!-- <section class="hero teaser">
<div class="container">
<div class="hero-body">
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/videos/teaser_vid.mp4"
type="video/mp4">
</video>
</div>
</div>
</section> -->
<!-- End teaser video -->
<!-- Paper abstract -->
<!-- <section class="section hero is-light">
<div class="container is-max-desktop">
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Abstract</h2>
<div class="content has-text-justified">
<p>
Many robot tasks require active or interactive exploration behavior in
order to be performed successfully. Such tasks are ubiquitous in embodied do-
mains, where agents must actively search for the information necessary for each
stage of a task, e.g., moving the head of the robot to find information relevant
to manipulation, or in multi-robot domains, where one scout robot may search
for the information that another robot needs to make informed decisions. We
identify these tasks with a new type of problem, factorized Contextual Markov
Decision Processes, and propose DISaM, a dual-policy solution composed of an
information-seeking policy that explores the environment to find the relevant con-
textual information and an information-receiving policy that exploits the context
to achieve the manipulation goal. This factorization allows us to train both poli-
cies separately, using the information-receiving one to provide reward to train the
information-seeking policy. At test time, the dual agent balances exploration and
exploitation based on the uncertainty the manipulation policy has on what the next
best action is. We demonstrate the capabilities of our dual policy solution in five
manipulation tasks that require information-seeking behaviors, both in simulation
and in the real-world, where DISaM significantly outperforms existing methods.
</p>
</div>
</div>
</div>
</div>
</section> -->
<!-- End paper abstract -->
<!-- Youtube video -->
<!-- <section class="hero is-small is-light">
<div class="hero-body">
<div class="container">
<h2 class="title is-3" style="text-align: center;">Video</h2>
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<div class="publication-video">
<iframe src="https://www.youtube.com/embed/7z6L_AGae6g" frameborder="0" allow="autoplay; encrypted-media" allowfullscreen></iframe>
</div>
</div>
</div>
</div>
</div>
</section> -->
<!-- End youtube video -->
<section class="section hero is-light">
<div class="container">
<h2 class="title is-2" style="text-align: center;">Overview</h2>
<div class="columns is-centered">
<div class="column is-four-fifths is-centered has-text-centered">
<div class="content has-text-justified">
<p>
Traditionally, task conditioned robot polcies assume access to all information about the task such as the
reward function being optimized but intelligent agents such as humans know how to look for important information
in their surroundings and take relevant actions based on the context. For example, when given the task of serving
a beverage, looking at the time of day can inform the agent what to serve.
</p>
</div>
<!-- Your image here -->
<div class="columns is-centered">
<div class="column is-four-fifths">
<video poster="" id="tree" autoplay controls muted loop height="60%">
<source src="static/videos/teaser_vid.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="section">
<div class="container">
<h2 class="title is-2" style="text-align: center;">DISaM Training</h2>
<div class="columns is-centered">
<div class="column is-four-fifths is-centered has-text-centered">
<div class="content has-text-justified">
<p>
In this work we factorize the problem of looking for information and acting as
information-seeking (IS) and information-receiving (IR) respectively, where we train
the IS agent to "look" for relevant task context and IR to act to complete the task.
Our method, DISaM (Dual Information-Seeking And Manipulation), splits the training into
two phases -- In Phase 1, we learn the IR policy that takes in ground-truth context
information and controls the movement of the robot. In Phase 2, we learn an IS policy
as well as an image encoder such that the context can be correctly reconstructed from
the camera observation. Once all parts are trained, together they create a system that
takes in image observations and controls both the robot and the camera.
</p>
</div>
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/videos/disam_training.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
</section>
<section class="section">
<div class="container">
<h2 class="title is-2" style="text-align: center;">DISaM Deployment</h2>
<div class="columns is-centered">
<div class="column is-four-fifths is-centered has-text-centered">
<div class="content has-text-justified">
<p>
During deployment, DISaM calculates the uncertainty of the IR policy over the next action by conditioning it on several contexts
generated with the Encoder. If the uncertainty of the IR policy is high (above a threshold) then information-seeking actions are taken by the IS
policy. When the correct context has been found by the IS policy, the IR uncertainty over the next action falls below the threshold
and DISaM takes the IR actions to complete the task.
</p>
</div>
</div>
</div>
<div class="container">
<div class="columns is-centered">
<div class="column is-three-fifths is-centered has-text-centered">
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/videos/deployment.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
</section>
<section class="section hero is-light">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column is-full-width">
<h2 class="title is-2" style="text-align: center;">Annotated Rollouts</h2>
<div class="content has-text-justified">
<p>
Following videos demonstrate how the control is switched between the Information-Seeking (IS) and Information-Receiving (IR)
agents. The text on top is semantic representation of the information that IR is uncertain about and we annotate the image
when IS is able to find that information.
</p>
</div>
<div class="grid-contrainer-one-no-box">
<div class="grid-item">
<div class="card-static-container">
<div class="card-container">
<div class="card-wide">
<div class="card-content">
<h3><strong>Task: </strong>Cooking a dish</h3>
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/videos/annotated_rollouts/cooking.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
<div class="card-container">
<div class="card-wide">
<div class="card-content">
<h3><strong>Task: </strong>Pick and place</h3>
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/videos/annotated_rollouts/walls.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
<div class="card-container">
<div class="card-wide">
<div class="card-content">
<h3><strong>Task: </strong>Assemble the correct peg</h3>
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/videos/annotated_rollouts/assembly.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
<div class="card-container">
<div class="card-wide">
<div class="card-content">
<h3><strong>Task: </strong>Choosing a beverage</h3>
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/videos/annotated_rollouts/clock.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
<div class="card-container">
<div class="card-wide">
<div class="card-content">
<h3><strong>Task: </strong>Serving a beverage</h3>
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/videos/annotated_rollouts/person.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
<div class="card-container">
<div class="card-wide">
<div class="card-content">
<h3><strong>Task: </strong>Pick fruit based on recipe</h3>
<video poster="" id="tree" autoplay controls muted loop height="100%">
<source src="static/videos/annotated_rollouts/button.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</div>
</section>
<section class="section">
<div class="container">
<h2 class="title is-2" style="text-align: center;">More Simulation Rollouts</h2>
<div class="columns is-centered">
<div class="column is-four-fifths is-centered has-text-centered">
<div class="content has-text-justified">
<p>
Below we provide more rollouts in the simulation environments to demonstrate the variety of behaviors IS policy learns.
The left frame corresponds to the the IS agent's observations and the right frame is a task visualization.
</p>
</div>
</div>
</div>
<!-- Cooking -->
<div class="columns is-centered">
<div class="column is-one-third">
<img src="static/images/task_gifs/cooking_1.gif" alt="cooking_1" width="90%"/>
</div>
<div class="column is-one-third">
<img src="static/images/task_gifs/cooking_2.gif" alt="cooking_2" width="90%"/>
<h4 class="title is-4" style="text-align: center;">Cooking</h3><br>
</div>
<div class="column is-one-third">
<img src="static/images/task_gifs/cooking_3.gif" alt="cooking_3" width="90%"/>
</div>
</div>
<!-- Walls -->
<div class="columns is-centered">
<div class="column is-one-third">
<img src="static/images/task_gifs/walls_1.gif" alt="walls_1" width="90%"/>
</div>
<div class="column is-one-third">
<img src="static/images/task_gifs/walls_2.gif" alt="walls_2" width="90%"/>
<h4 class="title is-4" style="text-align: center;">Walls</h3><br>
</div>
<div class="column is-one-third">
<img src="static/images/task_gifs/walls_3.gif" alt="walls_3" width="90%"/>
</div>
</div>
<!-- Assembly -->
<div class="columns is-centered">
<div class="column is-one-third">
<img src="static/images/task_gifs/assembly_1.gif" alt="assembly_1" width="90%"/>
</div>
<div class="column is-one-third">
<img src="static/images/task_gifs/assembly_2.gif" alt="assembly_2" width="90%"/>
<h4 class="title is-4" style="text-align: center;">Assembly</h3><br>
</div>
<div class="column is-one-third">
<img src="static/images/task_gifs/assembly_3.gif" alt="assembly_3" width="90%"/>
</div>
</div>
<!-- Real -->
<!-- <div class="columns is-centered">
<div class="column is-one-third">
<div class="columns is-centered">
<img src="static/images/task_gifs/real_button.gif" alt="button_1" width="75%"/>
</div>
<h4 class="title is-4" style="text-align: center;">Button</h3>
</div>
<div class="column is-one-third">
<img src="static/images/task_gifs/real_clock.gif" alt="clock_1" width="90%"/>
<h4 class="title is-4" style="text-align: center;">Teatime (Clock)</h3>
</div>
<div class="column is-one-third">
<img src="static/images/task_gifs/real_person.gif" alt="person_1" width="90%"/>
<h4 class="title is-4" style="text-align: center;">Teatime (Person)</h3>
</div>
</div> -->
</div>
</section>
<!-- <section class="section">
<div class="container" style="width: 70%;">
<h2 class="title is-2" style="text-align: center;">Autonomous Policy Rollouts</h2>
<div class="columns is-centered">
<div class="column">
<video poster="" id="video2" autoplay controls muted loop height="100%" style="border: 1px solid #bbb; border-radius: 10px; margin: 1.0%;">
<source src="static/videos/il/serve_bread_no_border.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<video poster="" id="video3" autoplay controls muted loop height="100%" style="border: 1px solid #bbb; border-radius: 10px; margin: 1.0%;">
<source src="static/videos/il/cover_table_no_border.mp4"
type="video/mp4">
</video>
</div>
<div class="column">
<video poster="" id="video3" autoplay controls muted loop height="100%" style="border: 1px solid #bbb; border-radius: 10px; margin: 1.0%;">
<source src="static/videos/il/slide_chair_no_border.mp4"
type="video/mp4">
</video>
</div>
</div>
</div>
</section> -->
<!--BibTex citation -->
<!-- <section class="section" id="BibTeX">
<div class="container is-max-desktop content">
<h2 class="title">BibTeX</h2>
<pre><code>
@article{dass2024telemoma,
title={TeleMoMa: A Modular and Versatile Teleoperation System for Mobile Manipulation},
author={Dass, Shivin and Ai, Wensi and Jiang, Yuqian and Singh, Samik and Hu, Jiaheng and Zhang, Ruohan and Stone, Peter and Abbatematteo, Ben and Martín-Martín, Roberto},
journal={arXiv preprint arXiv:2403.07869},
year={2024}
}
</code></pre>
</div>
</section> -->
<!--End BibTex citation -->
<footer class="footer">
<div class="container">
<div class="columns is-centered">
<div class="column is-8">
<div class="content">
<p>
This page was built using the <a href="https://github.com/eliahuhorwitz/Academic-project-page-template" target="_blank">Academic Project Page Template</a> which was adopted from the <a href="https://nerfies.github.io" target="_blank">Nerfies</a> project page.
You are free to borrow the of this website, we just ask that you link back to this page in the footer. <br> This website is licensed under a <a rel="license" href="http://creativecommons.org/licenses/by-sa/4.0/" target="_blank">Creative
Commons Attribution-ShareAlike 4.0 International License</a>.
</p>
</div>
</div>
</div>
</div>
</footer>
<!-- Statcounter tracking code -->
<!-- You can add a tracker to track page visits by creating an account at statcounter.com -->
<!-- End of Statcounter Code -->
</body>
</html>