-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathindex.html
More file actions
376 lines (331 loc) · 15.2 KB
/
index.html
File metadata and controls
376 lines (331 loc) · 15.2 KB
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
<!DOCTYPE html>
<html>
<style>
.has-text-red {
color: #bf5700;
}
.s {
color: #d14
}
.na {
color: #008080
}
.nc {
color: #445588;
font-weight: bold
}
.nl {
color: #990000;
font-weight: bold
}
</style>
<head>
<meta charset="utf-8">
<meta name="description"
content="Casper: Inferring Diverse Intents for Assistive Teleoperation with Vision Language Models">
<meta name="keywords" content="Assistive Teleoperation">
<meta name="viewport" content="width=device-width, initial-scale=1">
<title>Casper: Inferring Diverse Intents for Assistive Teleoperation with Vision Language Models</title>
<!-- Global site tag (gtag.js) - Google Analytics -->
<script async src="https://www.googletagmanager.com/gtag/js?id=G-PYVRSFMDRL"></script>
<script>
window.dataLayer = window.dataLayer || [];
function gtag() {
dataLayer.push(arguments);
}
gtag('js', new Date());
gtag('config', 'G-PYVRSFMDRL');
</script>
<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">
<link rel="icon" href="casper.png" type="image/png">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/3.5.1/jquery.min.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>
<nav class="navbar" role="navigation" aria-label="main navigation">
<div class="navbar-brand">
<a role="button" class="navbar-burger" aria-label="menu" aria-expanded="false">
<span aria-hidden="true"></span>
<span aria-hidden="true"></span>
<span aria-hidden="true"></span>
</a>
</div>
</nav>
<!-- <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">Casper: Inferring Diverse Intents for Assistive Teleoperation with Vision Language Models</h1>
<div class="is-size-5 publication-authors">
<span class="author-block" style="font-size: 25px;">Anonymous Authors</span>
</div>
<div class="item">
<img src="casper.png" alt="" style="width:10%"/>
</div>
</div>
</div>
</div>
</div>
</section> -->
<section class="hero">
<div class="hero-body">
<div class="container is-max-desktop">
<div class="columns is-centered">
<div class="column has-text-centered">
<!----- Title ----->
<h1 class="title is-2 publication-title">
Casper
<img src="casper.png" alt="casper icon"
style="height:1.5em; vertical-align:middle; transform: scaleX(-1); margin-left:-0.1em; margin-right:-0.1em;">
: Inferring Diverse Intents for Assistive Teleoperation with Vision Language Models
</h1>
<!----- Authors ----->
<p class="is-size-5 publication-authors">
<a href="#">Huihan Liu<sup>1</sup></a>,
<a href="#">Rutav Shah<sup>1</sup></a>,
<a href="#">Shuijing Liu<sup>1</sup></a>,
<a href="#">Jack Pittenger<sup>1</sup></a>,
<a href="#">Mingyo Seo<sup>1</sup></a>,<br>
<a href="#">Yuchen Cui<sup>2</sup></a>,
<a href="#">Yonatan Bisk<sup>3</sup></a>,
<a href="#">Roberto Martín-Martín<sup>1</sup></a>,
<a href="#">Yuke Zhu<sup>1</sup></a>
</p>
<!----- Institutions ----->
<p class="is-size-5 publication-authors">
<sup>1</sup> The University of Texas at Austin
<sup>2</sup> The University of California, Los Angeles
<br>
<sup>3</sup> Carnegie Mellon University
</p>
<br>
<div class="is-size-5 publication-authors">
<span class="author-block" style="color: #fa694e; font-size: 25px;">9th Conference on Robot Learning (CoRL 2025)</span>
</div>
<!----- Links ----->
<div class="publication-links" style="margin-top:1rem;">
<a href="https://arxiv.org/abs/2506.14727"
class="button is-rounded is-dark" style="margin:0 .5rem;">
<span class="icon"><i class="fas fa-file-pdf"></i></span><span>arXiv</span>
</a>
<a href="#"
class="button is-rounded is-dark" style="margin:0 .5rem;">
<span class="icon"><i class="fab fa-github"></i></span><span>Code (Coming Soon)</span>
</a>
</div>
<div class="is-flex is-justify-content-center is-align-items-center" style="gap:2rem; margin-top:0.5rem;">
<!-- UT Austin -->
<img src="images/ut-austin-logo.png"
alt="UT Austin logo"
style="max-height:60px; width:auto;">
<!-- UCLA -->
<img src="images/ucla-logo.png"
alt="UCLA logo"
style="max-height:90px; width:auto;">
<!-- CMU -->
<img src="images/cmu-logo.png"
alt="Carnegie Mellon University logo"
style="max-height:60px; width:auto;">
</div>
</div>
</div>
</div>
</div>
</section>
<section class="section">
<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">Casper Demo (Audio On!)</h2>
<div class="publication-video">
<!-- 16×9 responsive container -->
<figure class="image is-16by9">
<!-- Replace VIDEO_ID with the characters after v= in your YouTube URL -->
<iframe
class="has-ratio"
src="https://www.youtube.com/embed/ZRwjHh1t5kE?rel=0&showinfo=0"
title="Casper"
loading="lazy"
allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share"
allowfullscreen>
</iframe>
</figure>
</div>
<br>
</div>
</div>
<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>
Assistive teleoperation, where control is shared between a human and a robot, enables efficient and intuitive human-robot collaboration in diverse and unstructured environments. A central challenge in real-world assistive teleoperation is for the robot to infer a wide range of human intentions from user control inputs and to assist users with correct actions. Existing methods are either confined to simple, predefined scenarios or restricted to task-specific data distributions at training, limiting their support for real-world assistance. We introduce Casper, an assistive teleoperation system that leverages commonsense knowledge embedded in pre-trained visual language models (VLMs) for real-time intent inference and flexible skill execution. Casper incorporates an open-world perception module for a generalized understanding of novel objects and scenes, a VLM-powered intent inference mechanism that leverages commonsense reasoning to interpret snippets of teleoperated user input, and a skill library that expands the scope of prior assistive teleoperation systems to support diverse, long-horizon mobile manipulation tasks. Extensive empirical evaluation, including human studies and system ablations, demonstrates that Casper improves task performance, reduces human cognitive load, and achieves higher user satisfaction than direct teleoperation and assistive teleoperation baselines.
</p>
<br>
</div>
</div>
</div>
<!-- <div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Demo</h2>
<div class="publication-video">
<video muted autoplay controls width="100%">
<source src="videos/casper_demo.mp4" type="video/mp4">
</video>
</div>
</div>
</div> -->
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Overview</h2>
<div class="content has-text-justified">
<p>
Casper infers user intents and offers help when confident. Given user teleoperation input, Casper uses VLMs to predict human intent using commonsense reasoning. Upon user confirmation, Casper performs autonomous execution to fulfill the intent using a skill library. Casper's background reasoning runs in parallel with foreground human control to minimize disruption.
</p>
</div>
<div class="publication-video">
<video muted autoplay controls width="100%">
<source src="videos/overview.mp4" type="video/mp4">
</video>
</div>
<br>
<br>
</div>
</div>
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Method</h2>
<div class="content has-text-justified">
<p>
Casper has VLM-based intent inference run in parallel with human teleoperation. Casper generates task candidates from observations and infers intent from user inputs among the task candidates, repeating until predictions are self-consistent. Once confirmed by the user, Casper executes the corresponding skill with estimated parameters.
</p>
</div>
<div class="publication-video">
<video muted autoplay controls width="100%">
<source src="videos/method.mp4" type="video/mp4">
</video>
</div>
<br>
<br>
</div>
</div>
<div class="columns is-centered has-text-centered">
<div class="column is-four-fifths">
<h2 class="title is-3">Experiments</h2>
<div class="content has-text-justified">
<p>
We conduct experiments on 3 multi-step mobile manipulation tasks: Toy, Shelf and Door. At each step, the robot disambiguates user intent among multiple plausible goals, selecting the correct one based on user inputs and visual context.
</p>
</div>
<div class="item">
<!-- Your image here -->
<img src="videos/tasks.png" alt=""/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<br>
<div class="content has-text-justified">
<p>
We conducted an IRB-approved user study with N=13 participants, all of whom gave informed consent. Participants completed a practice session before completing each method in randomized order. After each, they answered user satisfaction and NASA-TLX questionnaires.
</p>
</div>
<div class="item">
<!-- Your image here -->
<img src="videos/human_study.png" alt="" style="width:100%; height:auto;"/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<br>
<div class="content has-text-justified">
<p>
<strong>User study: user workload and user satisfaction.</strong> Casper consistently outperforms the baselines in terms of user workload (left) and user satisfaction (right) with statistical significance. Note that for user satisfaction scores, ''assist helpfully'' and ''correct intent'' are not applicable to Full Teleop.
</p>
</div>
<div class="item">
<!-- Your image here -->
<img src="videos/results-user.png" alt=""/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<br>
<div class="content has-text-justified">
<p>
<strong>User study: task success rate and completion time.</strong> Casper outperforms baselines in both task success and completion time.
</p>
</div>
<div class="item">
<!-- Your image here -->
<img src="videos/results-success.png" alt="" style="width:50%; height:auto;"/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<br>
<div class="content has-text-justified">
<p>
<strong>Quantitative results from unit testing and ablation studies:</strong>
<br>
<strong>Left</strong>: Casper outperforms all baselines in intent inference success rate. Note that no STD is reported for deterministic baselines. The ablation of Casper vs. Capser - No Visual Prompting (VP) highlights the benefit of visual prompting.
<br>
<strong>Middle</strong>: Success rates improve with longer teleoperation history.
<br>
<strong>Right</strong>: Removing confidence estimation increases false prediction rates across all history lengths.
<br>
</p>
</div>
<div class="item">
<!-- Your image here -->
<img src="videos/ablations.png" alt=""/>
<h2 class="subtitle has-text-centered">
</h2>
</div>
<br>
</div>
</div>
<!--
<div class="column is-four-fifths">
<h2 class="title is-3">Overview</h2>
<div class="content has-text-justified">
<p>
Casper infers user intents and offers help when confident. Given user teleoperation input, Casper uses VLMs to predict human intent using commonsense reasoning. Upon user confirmation, Casper performs autonomous execution to fulfill the intent using a skill library. Casper's background reasoning runs in parallel with foreground human control to minimize disruption.
</p>
<br>
</div>
<div class="publication-video">
<video muted autoplay controls width="100%">
<source src="videos/overview.mp4" type="video/mp4">
</video>
</div>
<br>
<br>
</div>
</div> -->
<div class="columns is-centered">
<div class="column is-four-fifths">
<h2 class="title is-3 has-text-centered">BibTeX</h2>
<pre><code>
@misc{liu2025casperinferringdiverseintents,
title={Casper: Inferring Diverse Intents for Assistive Teleoperation with Vision Language Models},
author={Huihan Liu and Rutav Shah and Shuijing Liu and Jack Pittenger and Mingyo Seo and Yuchen Cui and Yonatan Bisk and Roberto Martín-Martín and Yuke Zhu},
year={2025},
eprint={2506.14727},
archivePrefix={arXiv},
primaryClass={cs.RO},
url={https://arxiv.org/abs/2506.14727},
}
</code></pre>
</div>
</div>
</body>
</html>