PageSourceSearch

https://haonan-duan.github.io/

html haonan-duan.github.io collected 2026-10-03 09:16:38 UTC 15,960 bytes, 313 lines download raw bytes

1<!DOCTYPE html>
2<html lang="en">
3<head>
4  <meta charset="utf-8">
5  <meta name="viewport" content="width=device-width, initial-scale=1">
6  
6<script>document.documentElement.className += ' js';</script>
6
7
8  <meta name="description" content="🤖🍣✈️🚃🏎️🌁🍦">
9  <meta name="author" content="Haonan Duan">
10  <title>Haonan Duan</title>
11
12  <link rel="canonical" href="https://haonan-duan.github.io/">
13  <meta property="og:type" content="profile">
14  <meta property="og:title" content="Haonan Duan">
15  <meta property="og:description" content="🤖🍣✈️🚃🏎️🌁🍦">
16  <meta property="og:image" content="https://haonan-duan.github.io/images/haonan.png">
17  <meta property="og:url" content="https://haonan-duan.github.io/">
18  <meta name="twitter:card" content="summary">
19
20  <link rel="apple-touch-icon" sizes="180x180" href="/images/apple-touch-icon.png">
21  <link rel="icon" type="image/png" sizes="32x32" href="/images/favicon-32x32.png">
22  <link rel="icon" type="image/png" sizes="16x16" href="/images/favicon-16x16.png">
23  <meta name="theme-color" content="#fbf9fe">
24
25  <link rel="stylesheet" href="/assets/css/main.css">
26</head>
27
28<body data-author-email="[email protected]">
29  <div class="blobs" aria-hidden="true"><span></span><span></span><span></span><span></span></div>
30
31  <div class="shell">
32    <aside class="sidebar" aria-label="profile">
33  <div class="card card--pink profile">
34    <div class="profile__portrait">
35      <img class="profile__avatar" src="/images/haonan.png" alt="Haonan Duan">
36    </div>
37
38    <h1 class="profile__name">Haonan Duan</h1>
39    <p class="profile__bio">Robotics Learner</p>
40    <p class="profile__loc">
41      <svg viewBox="0 0 16 16" width="12" height="12" aria-hidden="true"><path fill="currentColor" d="M8 0C4.7 0 2 2.7 2 6c0 4.5 6 10 6 10s6-5.5 6-10c0-3.3-2.7-6-6-6zm0 8.2A2.2 2.2 0 1 1 8 3.8a2.2 2.2 0 0 1 0 4.4z"/></svg>
42      Shanghai, China
43    </p>
44
45    <nav class="profile__nav" aria-label="sections">
46      <a href="#about"        data-dot="coral"><span></span>about</a>
47      <a href="#news"         data-dot="amber"><span></span>news</a>
48      <a href="#publications" data-dot="violet"><span></span>publications</a>
49      <a href="#education"    data-dot="sky"><span></span>education</a>
50      <a href="#experience"   data-dot="mint"><span></span>experience</a>
51    </nav>
52
53    <ul class="profile__links">
54      <li><a href="mailto:[email protected]" aria-label="email" title="email">
55        <svg viewBox="0 0 16 16" width="17" height="17"><path fill="currentColor" d="M1.5 3h13c.3 0 .5.2.5.5v9c0 .3-.2.5-.5.5h-13a.5.5 0 0 1-.5-.5v-9c0-.3.2-.5.5-.5zm.7 1.3v.6L8 9l5.8-4.1v-.6H2.2zm11.6 1.8L8 10.3 2.2 6.1v6.2h11.6V6.1z"/></svg>
56      </a></li>
57      <li><a href="https://scholar.google.com/citations?user=4zO4UlcAAAAJ" target="_blank" rel="noopener" aria-label="google scholar" title="google scholar">
58        <svg viewBox="0 0 16 16" width="17" height="17"><path fill="currentColor" d="M8 0L0 5.5 8 11l6.5-4.5V11h1V5.5L8 0zM4 9.3v2.2c0 1.2 1.8 2.5 4 2.5s4-1.3 4-2.5V9.3l-4 2.7-4-2.7z"/></svg>
59      </a></li>
60      <li><a href="https://github.com/haonan-duan" target="_blank" rel="noopener" aria-label="github" title="github">
61        <svg viewBox="0 0 16 16" width="17" height="17"><path fill="currentColor" d="M8 0C3.6 0 0 3.6 0 8c0 3.5 2.3 6.5 5.5 7.6.4.1.5-.2.5-.4v-1.4c-2.2.5-2.7-1-2.7-1-.4-1-.9-1.2-.9-1.2-.7-.5.1-.5.1-.5.8.1 1.2.8 1.2.8.7 1.2 1.9.9 2.4.7.1-.5.3-.9.5-1.1-1.8-.2-3.6-.9-3.6-4 0-.9.3-1.6.8-2.1-.1-.2-.4-1 .1-2.1 0 0 .7-.2 2.2.8.6-.2 1.3-.3 2-.3s1.4.1 2 .3c1.5-1 2.2-.8 2.2-.8.4 1.1.2 1.9.1 2.1.5.5.8 1.2.8 2.1 0 3.1-1.9 3.7-3.6 3.9.3.3.6.8.6 1.6v2.3c0 .2.1.5.6.4C13.7 14.5 16 11.5 16 8c0-4.4-3.6-8-8-8z"/></svg>
62      </a></li>
63      <li><a href="https://linkedin.com/in/haonan-duan" target="_blank" rel="noopener" aria-label="linkedin" title="linkedin">
64        <svg viewBox="0 0 16 16" width="17" height="17"><path fill="currentColor" d="M13.6 0H2.4A2.4 2.4 0 0 0 0 2.4v11.2A2.4 2.4 0 0 0 2.4 16h11.2a2.4 2.4 0 0 0 2.4-2.4V2.4A2.4 2.4 0 0 0 13.6 0zM4.8 13.3H2.7V6.1h2.1v7.2zM3.7 5.2a1.2 1.2 0 1 1 0-2.5 1.2 1.2 0 0 1 0 2.5zm9.6 8.1h-2.1V9.8c0-.9 0-2-1.2-2s-1.4.9-1.4 1.9v3.6H6.5V6.1h2.1v1c.3-.6 1-1.2 2.1-1.2 2.2 0 2.6 1.5 2.6 3.4v4z"/></svg>
65      </a></li>
66    </ul>
67
68    <p class="profile__hint">press <kbd>⌘K</kbd> to jump</p>
69  </div>
70</aside>
71
72    <main class="main" role="main">
73      <section class="card card--coral reveal" id="about">
74  <header class="card__head">
75    <h2 class="card__title"><span class="card__icon">👋</span>About</h2>
76  </header>
77<div>
78
79    <p>
79I am a researcher at <a href="https://research.nvidia.com/labs/gear/">NVIDIA GEAR</a>, working under the guidance of Dr. <a href="https://jimfan.me/">Jim Fan</a> and Prof. <a href="https://yukezhu.me/">Yuke Zhu</a>. My research sits at the intersection of Multimodal Large Language Models and Robot Learning, with a specific focus on building foundation models for robotic perception and manipulation.</p>
80
81    <p>Previously, I worked on robot foundation models at Xiaomi. Before that, I was a researcher at SenseTime, where I worked on Vision-Language-Action models across pre-training, post-training, and their applications in multi-view generalization and human-robot interaction. I also contributed to video understanding, particularly agent-based reinforcement learning for long-horizon video reasoning.</p>
82
83    <p>I received my Ph.D. from the Institute of Automation, Chinese Academy of Sciences in 2024. My doctoral research focused on robotic dexterous grasping and human–robot interaction.</p>
84
85  </div>
86</section>
87
88<section class="card card--amber reveal" id="news">
89  <header class="card__head">
90    <h2 class="card__title"><span class="card__icon">✨</span>News</h2>
91    <span class="card__chip">7 updates</span>
92  </header>
93  <ul class="feed">
94    <li class="feed__item">
95      <span class="feed__date">2025.11</span>
96      <span class="feed__body">One <a href="https://arxiv.org/abs/2508.13103v1">paper</a> got accepted by AAAI 2026.</span>
97    </li>
98    <li class="feed__item">
99      <span class="feed__date">2025.06</span>
100      <span class="feed__body">One <a href="https://arxiv.org/abs/2503.19757">paper</a> got accepted by ICCV 2025.</span>
101    </li>
102    <li class="feed__item">
103      <span class="feed__date">2024.11</span>
104      <span class="feed__body">One <a href="https://ieeexplore.ieee.org/document/10770603">paper</a> got accepted by IEEE Transactions on Robotics (T-RO).</span>
105    </li>
106    <li class="feed__item">
107      <span class="feed__date">2024.06</span>
108      <span class="feed__body">One <a href="https://ieeexplore.ieee.org/document/10577462">paper</a> got accepted by IEEE Transactions on Robotics (T-RO).</span>
109    </li>
110    <li class="feed__item">
111      <span class="feed__date">2024.05</span>
112      <span class="feed__body">Finished my Ph.D. final defense! What an unforgettable journey!</span>
113    </li>
114    <li class="feed__item">
115      <span class="feed__date">2024.01</span>
116      <span class="feed__body">One <a href="https://ieeexplore.ieee.org/document/10610646">paper</a> got accepted by ICRA 2024.</span>
117    </li>
118    <li class="feed__item">
119      <span class="feed__date">2023.01</span>
120      <span class="feed__body">Started internship at NIO Autonomous Driving.</span>
121    </li>
122  </ul>
123</section>
124
125<section class="card card--violet reveal" id="publications">
126  <header class="card__head">
127    <h2 class="card__title"><span class="card__icon">📄</span>Publications</h2>
128    <span class="card__chip">6 selected</span>
129  </header>
130  <p class="card__lede">
131    See <a href="https://scholar.google.com/citations?user=4zO4UlcAAAAJ">Google Scholar</a> for the full list.
132  </p>
133  <div class="papers"><article class="paper">
134      <a class="paper__figure" href="https://arxiv.org/abs/2508.13103v1" aria-hidden="true" tabindex="-1">
135        <img src="/images/OC-VLA.png" alt="" loading="lazy" />
136      </a>
137      <div class="paper__body">
138        <h3 class="paper__title">
139          <a href="https://arxiv.org/abs/2508.13103v1">Grounding Actions in Camera Space: Observation-Centric Vision-Language-Action Policy</a>
140        </h3>
141        <p class="paper__authors">Tianyi Zhang<sup>*</sup>, <strong>Haonan Duan</strong><sup>*</sup>, Haoran Hao, Yu Qiao, Jifeng Dai, Zhi Hou</p>
142        <p class="paper__meta">
143          <span class="paper__venue" title="AAAI, 2026">AAAI 2026</span>
144            <a class="paper__link" href="https://arxiv.org/abs/2508.13103v1">Paper</a>
145        </p>
146      </div>
147    </article>
148<article class="paper">
149      <a class="paper__figure" href="https://arxiv.org/abs/2503.19757" aria-hidden="true" tabindex="-1">
150        <img src="/images/Dita.png" alt="" loading="lazy" />
151      </a>
152      <div class="paper__body">
153        <h3 class="paper__title">
154          <a href="https://arxiv.org/abs/2503.19757">Dita: Scaling Diffusion Transformer for Generalist Vision-Language-Action Policy</a>
155        </h3>
156        <p class="paper__authors">Zhi Hou<sup>*</sup>, Tianyi Zhang<sup>*</sup>, Yuwen Xiong, <strong>Haonan Duan</strong>, Hengjun Pu, Ronglei Tong, Chengyang Zhao, Xizhou Zhu, Yu Qiao, Jifeng Dai, Yuntao Chen</p>
157        <p class="paper__meta">
158          <span class="paper__venue" title="ICCV, 2025">ICCV 2025</span>
159            <a class="paper__link" href="https://robodita.github.io/">Project</a>
160            <a class="paper__link" href="https://arxiv.org/abs/2503.19757">Paper</a>
161            <a class="paper__link" href="https://www.bilibili.com/video/BV1j5QBYzEkR">Bilibili</a>
162            <a class="paper__link" href="https://www.youtube.com/watch?v=pBLcmeg0mdY">YouTube</a>
163            <a class="paper__link" href="https://zhuanlan.zhihu.com/p/31684414907">知乎</a>
164        </p>
165      </div>
166    </article>
167<article class="paper">
168      <a class="paper__figure" href="https://ieeexplore.ieee.org/document/10770603" aria-hidden="true" tabindex="-1">
169        <img src="/images/TRO-D.png" alt="" loading="lazy" />
170      </a>
171      <div class="paper__body">
172        <h3 class="paper__title">
173          <a href="https://ieeexplore.ieee.org/document/10770603">Reactive Human-to-Robot Dexterous Handovers for Anthropomorphic Hand</a>
174        </h3>
175        <p class="paper__authors"><strong>Haonan Duan</strong>, Peng Wang, Yifan Yang, Daheng Li, Wei Wei, Yongkang Luo, Guoqiang Deng</p>
176        <p class="paper__meta">
177          <span class="paper__venue" title="IEEE Transactions on Robotics, 2024">T-RO 2024</span>
178            <a class="paper__link" href="https://ieeexplore.ieee.org/document/10770603">Paper</a>
179        </p>
180      </div>
181    </article>
182<article class="paper">
183      <a class="paper__figure" href="https://ieeexplore.ieee.org/document/10577462" aria-hidden="true" tabindex="-1">
184        <img src="/images/TRO-W.png" alt="" loading="lazy" />
185      </a>
186      <div class="paper__body">
187        <h3 class="paper__title">
188          <a href="https://ieeexplore.ieee.org/document/10577462">Learning Human-like Functional Grasping for Multi-finger Hands from Few Demonstrations</a>
189        </h3>
190        <p class="paper__authors">Wei Wei, Peng Wang, Sizhe Wang, Yongkang Luo, Wanyi Li, Daheng Li, Yayu Huang, <strong>Haonan Duan</strong></p>
191        <p class="paper__meta">
192          <span class="paper__venue" title="IEEE Transactions on Robotics, 2024">T-RO 2024</span>
193            <a class="paper__link" href="https://v-wewei.github.io/sr_dexgrasp/">Project</a>
194            <a class="paper__link" href="https://www.bilibili.com/video/BV1Pe411n7w6/?spm_id_from=888.80997.embed_other.whitelist&amp;t=18.087712&amp;bvid=BV1Pe411n7w6&amp;vd_source=1f3c77957407e0c8a5f2cd9dd9baede0">Video</a>
195        </p>
196      </div>
197    </article>
198<article class="paper">
199      <a class="paper__figure" href="https://ieeexplore.ieee.org/document/10610646" aria-hidden="true" tabindex="-1">
200        <img src="/images/ICRA2024.png" alt="" loading="lazy" />
201      </a>
202      <div class="paper__body">
203        <h3 class="paper__title">
204          <a href="https://ieeexplore.ieee.org/document/10610646">Learning Realistic and Reasonable Grasps for Anthropomorphic Hand in Cluttered Scenes</a>
205        </h3>
206        <p class="paper__authors"><strong>Haonan Duan</strong>, Yiming Li, Daheng Li, Wei Wei, Yayu Huang, Peng Wang</p>
207        <p class="paper__meta">
208          <span class="paper__venue" title="ICRA, 2024">ICRA 2024</span>
209            <a class="paper__link" href="https://ieeexplore.ieee.org/document/10610646">Paper</a>
210        </p>
211      </div>
212    </article>
213<article class="paper">
214      <a class="paper__figure" href="https://ieeexplore.ieee.org/document/9870872" aria-hidden="true" tabindex="-1">
215        <img src="/images/TCDS2022.png" alt="" loading="lazy" />
216      </a>
217      <div class="paper__body">
218        <h3 class="paper__title">
219          <a href="https://ieeexplore.ieee.org/document/9870872">Learning Human-to-Robot Dexterous Handovers for Anthropomorphic Hand</a>
220        </h3>
221        <p class="paper__authors"><strong>Haonan Duan</strong>, Peng Wang, Yiming Li, Daheng Li, Wei Wei</p>
222        <p class="paper__meta">
223          <span class="paper__venue" title="IEEE Transactions on Cognitive and Developmental Systems, 2022">T-CDS 2022</span>
224            <a class="paper__link" href="https://ieeexplore.ieee.org/document/9870872">Paper</a>
225        </p>
226      </div>
227    </article>
228
229  </div>
230</section>
231
232<section class="card card--sky reveal" id="education">
233  <header class="card__head">
234    <h2 class="card__title"><span class="card__icon">🎓</span>Education</h2>
235  </header>
236  <ul class="feed">
237    <li class="feed__item">
238      <span class="feed__date">2021 – 2024</span>
239      <span class="feed__body">
240        <strong>Ph.D.</strong>, Institute of Automation, Chinese Academy of Sciences
241        <span class="feed__meta">Beijing, China</span>
242      </span>
243    </li>
244    <li class="feed__item">
245      <span class="feed__date">2019 – 2021</span>
246      <span class="feed__body">
247        <strong>M.Sc.</strong>, University of Pittsburgh
248        <span class="feed__meta">Pittsburgh, PA, USA</span>
249      </span>
250    </li>
251  </ul>
252</section>
253
254<section class="card card--mint reveal" id="experience">
255  <header class="card__head">
256    <h2 class="card__title"><span class="card__icon">💼</span>Experience</h2>
257  </header>
258  <ul class="feed">
259    <li class="feed__item feed__item--current">
260      <span class="feed__date">2025.12 – now</span>
261      <span class="feed__body">
262        <strong>Researcher</strong>
263        <span class="feed__meta">NVIDIA GEAR</span>
264      </span>
265    </li>
266    <li class="feed__item">
267      <span class="feed__date">2025.09 – 2025.12</span>
268      <span class="feed__body">
269        <strong>Senior Researcher</strong>
270        <span class="feed__meta">Xiaomi Robot Foundation Models</span>
271      </span>
272    </li>
273    <li class="feed__item">
274      <span class="feed__date">2024.07 – 2025.09</span>
275      <span class="feed__body">
276        <strong>Senior Researcher</strong>
277        <span class="feed__meta">SenseTime Research</span>
278      </span>
279    </li>
280    <li class="feed__item">
281      <span class="feed__date">2023.01 – 2023.06</span>
282      <span class="feed__body">
283        <strong>Algorithm Engineer Intern</strong>
284        <span class="feed__meta">NIO Autonomous Driving</span>
285      </span>
286    </li>
287  </ul>
288</section>
289
290    </main>
291  </div>
292
293  <footer class="footer">
294    <p class="footer__count">visitors · <b id="visitor-count" aria-live="polite">—</b></p>
295    <p>built with <a href="https://claude.com/claude-code">claude code</a></p>
296  </footer>
297
298  
299
300<!-- Global site tag (gtag.js) - Google Analytics -->
301<script async src="https://www.googletagmanager.com/gtag/js?id="></script>
vendor: 1 bytes, line 301
301
302<script>
303  window.dataLayer = window.dataLayer || [];
304  function gtag(){dataLayer.push(arguments);}
305  gtag('js', new Date());
306
307  gtag('config', "");
308</script>
308
309
310
311  
311<script src="/assets/js/site.js" defer></script>
311
312</body>
313</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.