forked from matcha-model/matcha-model.github.io
-
Notifications
You must be signed in to change notification settings - Fork 0
/
index.html
226 lines (210 loc) · 10.8 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
<!DOCTYPE html>
<html>
<head lang="en">
<meta charset="UTF-8">
<meta http-equiv="x-ua-compatible" content="ie=edge">
<title>Matcha agent - Chat with the Environment: Interactive Multimodal Perception using Large Language Models
</title>
<meta name="description"
content="Chat with the Environment: Interactive Multimodal Perception using Large Language Models">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.5/css/bootstrap.min.css">
<link rel="stylesheet" href="https://maxcdn.bootstrapcdn.com/font-awesome/4.4.0/css/font-awesome.min.css">
<link rel="stylesheet" href="https://cdnjs.cloudflare.com/ajax/libs/codemirror/5.8.0/codemirror.min.css">
<link rel="stylesheet" href="css/app.css">
<link rel="stylesheet" href="css/bootstrap.min.css">
<link rel="stylesheet" href="css/background.css">
<link rel="stylesheet" href="https://cdn.jsdelivr.net/npm/locomotive-scroll@4.1.3/dist/locomotive-scroll.css">
<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.11.3/jquery.min.js"></script>
<script src="https://maxcdn.bootstrapcdn.com/bootstrap/3.3.5/js/bootstrap.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/codemirror/5.8.0/codemirror.min.js"></script>
<script src="https://cdnjs.cloudflare.com/ajax/libs/clipboard.js/1.5.3/clipboard.min.js"></script>
<script src="js/app.js"></script>
<!-- <script src="https://cdn.jsdelivr.net/npm/ukiyojs@4.0.3/dist/ukiyo.umd.min.js"> </script> -->
<style>
.nav-pills {
position: relative;
display: inline;
}
.imtip {
position: absolute;
top: 0;
left: 0;
}
</style>
</head>
<body>
<div id="parallax-world-of-ugg">
<section>
<div class="title">
<h3>IROS 2023, Detroit</h3>
<h1>Matcha Agent</h1>
</div>
</section>
<section>
<div class="parallax-one">
<h2>Chat with the Environment:
Interactive Multimodal Perception using Large Language Models
</h2>
</div>
</section>
</br>
</br>
<div class="container" id="main">
<div class="row">
<p class="col-md-12 text-center">
<font size="+3">Chat with the Environment: </br>
Interactive Multimodal Perception using Large Language Models
</font>
</br>
</p>
<div class="col-md-12 text-center">
(IROS 2023, Detroit – IEEE/RSJ International Conference on Intelligent Robots)
</div>
</div>
<div class="row">
<div class="col-md-12 text-center">
<ul class="list-inline">
<br>
<li><a href="https://xf-zhao.github.io">Xufeng Zhao</a> </li>
<li><a href="https://mengdi-li.github.io/">Mengdi Li</a></li>
<li><a href="https://www.inf.uni-hamburg.de/en/inst/ab/wtm/people/weber.html">Cornelius
Weber</a></li>
<li><a href="https://www.mbhafez.com/">Muhammad Burhan Hafez</a></li>
<li><a href="https://www.inf.uni-hamburg.de/en/inst/ab/wtm/people/wermter.html">Stefan
Wermter</a></li>
<br><br>
<a href="https://www.inf.uni-hamburg.de/en/inst/ab/wtm.html">
<image src="img/new-wtm-logo-white-150x150.jpg" height="40px"> Knowledge Technology Group, University of Hamburg
</a>
</ul>
</div>
</div>
<div class="row">
<div class="col-md-4 col-md-offset-4 text-center">
<ul class="nav nav-pills nav-justified">
<li>
<a href="https://arxiv.org/abs/2303.08268">
<image src="img/paper_small.png" height="60px">
<h4><strong>Paper</strong></h4>
</a>
</li>
<li>
<a href="img/matcha-poster.pdf">
<image src="img/poster_small.png" height="60px">
<h4><strong>Poster</strong></h4>
</a>
</li>
<li>
<a href="img/matcha-agent-slides.pdf">
<image src="img/slide.png" height="60px">
<!-- <image src="img/new.png" height="20px" class="imtip"> -->
<h4><strong>Slides</strong></h4>
</a>
</li>
<li>
<a href="https://github.com/xf-zhao/Matcha">
<image src="img/github.png" height="60px">
<h4><strong>Code</strong></h4>
</a>
</li>
</ul>
</div>
</div>
<div class="row">
<div class="col-md-8 text-left col-md-offset-2">
<h4>
🔔 <b>News</b>
</h4>
<li> <a style="color:gray;">[2023-09-29]</a> Add IORS 2023 slides.</li>
<li> <a style="color:gray;">[2023-09-29]</a> Refactored and open-sourced all the <b>codes</b> for Matcha-Agent!</li>
<li> <a style="color:gray;">[2023-09-20]</a> Add IROS 2023 <b>poster</b>. </li>
</div>
</div>
</br>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h4><b>
📜 Abstract </b>
</h4>
<p class="text-justify">
Programming robot behavior in a complex world faces challenges on multiple levels, from dextrous
low-level skills to high-level planning and reasoning.
Recent pre-trained Large Language Models (LLMs) have shown remarkable reasoning ability in
few-shot robotic planning.
However, it remains challenging to ground LLMs in multimodal sensory input and continuous action
output, while enabling a robot to interact with its environment and acquire novel information as
its policies unfold.
We develop a robot interaction scenario with a partially observable state, which necessitates a
robot to decide on a range of epistemic actions in order to sample sensory information among
multiple modalities, before being able to execute the task correctly.
<b>Matcha</b> (<b>M</b>ultimod<b>a</b>l environmen<b>t</b> <b>cha</b>tting) <b>agent</b>, an
interactive perception framework, is therefore proposed with an LLM as its backbone, whose
ability is exploited to instruct epistemic actions and to reason over the resulting multimodal
sensations (vision, sound, haptics, proprioception), as well as to plan an entire task execution
based on the interactively acquired information.
Our study demonstrates that LLMs can provide high-level planning and reasoning skills and
control interactive robot behavior in a multimodal environment, while multimodal modules with
the context of the environmental state help ground the LLMs and extend their processing ability.
The project website can be found at <a href="">https://matcha-agent.github.io</a>.
</p>
</div>
</div>
<div class="row">
<div class="col-md-8 text-left col-md-offset-2">
<h4>
🌄 <b>Video demo</b>
</h4>
<li>
<b>Matcha agent</b> manipulates objects with different sound, weights and haptics to determine
their materials.
</li>
<li>
<a href="https://arxiv.org/abs/2305.08528">NICOL</a> robot from <a
href="https://www.inf.uni-hamburg.de/en/inst/ab/wtm.html">Knowledge Technology Group,
University of Hamburg</a>.
</li>
<li>
In CoppeliaSim simulator.
</li>
<li>
Please <b>turn on</b> your speaker to hear the sound!
</li>
</br>
<div class="text-center">
<div style="position:relative;padding-top:56.25%;">
<iframe width="560" height="315" src="https://www.youtube.com/embed/rMMeMTWmT0k"
allowfullscreen style="position:absolute;top:0;left:0;width:100%;height:100%;"></iframe>
</div>
</div>
</br>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
<h4><b>
Citation</b>
</h4> <a href="https://arxiv.org/abs/2303.08268">[arxiv version]</a>
<div class="form-group col-md-10 col-md-offset-1">
<textarea id="bibtex" class="form-control" readonly>
@article{zhao2023chat,
title={Chat with the environment: Interactive multimodal perception using large language models},
author={Zhao, Xufeng and Li, Mengdi and Weber, Cornelius and Hafez, Muhammad Burhan and Wermter, Stefan},
journal={arXiv preprint arXiv:2303.08268},
year={2023}
}
</textarea>
</div>
</div>
</div>
<div class="row">
<div class="col-md-8 col-md-offset-2">
This website template borrowed from <a href="https://say-can.github.io/">say-can</a>,
but adapted with background animation. Feel free to fork <a
href="https://github.com/matcha-agent/matcha-agent.github.io">this template</a>.
</div>
</div>
</div>
</br>
</body>
</html>