index.html

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8">
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  <title>Embodied AI (EAI) Lab</title>
  <base target="_blank">
  <link href="https://fonts.googleapis.com/css2?family=Libre+Baskerville:wght@400;700&family=Source+Sans+3:wght@400;500;600&display=swap" rel="stylesheet">
  <style>
    :root {
      --pub-link-color: #2563eb;
      --pub-link-underline: #93c5fd;
      --pub-sup-color: #52525b;
    }

    * {
      margin: 0;
      padding: 0;
      box-sizing: border-box;
    }
    
    body {
      font-family: 'Libre Baskerville', Georgia, serif;
      color: #1f2937;
      line-height: 1.7;
      background: #fff;
      font-size: 18px;
    }
    
    h1, h2 {
      font-family: 'Libre Baskerville', Georgia, serif;
      font-weight: 700;
    }
    
    a {
      text-decoration: none;
      color: #2563eb;
    }
    
    a:hover {
      text-decoration: underline;
    }
    
    .container {
      max-width: 1000px;
      margin: 0 auto;
      padding: 3rem 2rem;
    }
    
    @media (max-width: 768px) {
      .container {
        padding: 2rem 1.25rem;
      }
      body {
        font-size: 16px;
      }
    }
    
    /* Header */
    .header {
      margin-bottom: 2rem;
    }
    
    .header h1 {
      font-size: 2rem;
      color: #1f2937;
      margin-bottom: 0.75rem;
      line-height: 1.3;
    }
    
    @media (max-width: 768px) {
      .header h1 {
        font-size: 1.5rem;
      }
    }
    
    .header-line {
      height: 3px;
      background: #1e3a5f;
    }
    
    /* Description */
    .description {
      margin-bottom: 3rem;
      color: #374151;
    }
    
    .description p {
      margin-bottom: 1rem;
    }
    
    /* Section */
    section {
      margin-bottom: 3rem;
    }
    
    section h2 {
      font-size: 1.5rem;
      color: #1f2937;
      margin-bottom: 0.5rem;
    }

    @media (max-width: 768px) {
      section h2 {
        font-size: 1.25rem;
      }
    }
    
    .section-line {
      height: 2px;
      background: #1e3a5f;
      margin-bottom: 1.5rem;
    }

    .section-note {
      font-size: 0.7em;
      font-weight: 400;
      color: #1e3a5f;
      letter-spacing: 0.01em;
    }

    .section-note sup {
      font-size: 0.8em;
      vertical-align: super;
      line-height: 0;
      color: var(--pub-sup-color);
    }
    
    /* Publications */
    .publications ul {
      list-style: disc;
      padding-left: 1.5rem;
    }
    
    .publications li {
      margin-bottom: 1.75rem;
    }
    
    .pub-title {
      color: #111111;
      font-weight: 700;
      font-size: 1.05em;
      display: block;
      margin-bottom: 0.25rem;
    }
    
    .pub-authors {
      color: #1f1f1f;
      margin-bottom: 0.15rem;
    }

    .pub-authors strong {
      color: inherit;
      font-weight: inherit;
    }
    
    .pub-authors sup {
      font-size: 0.7em;
      margin-left: 1px;
      color: var(--pub-sup-color);
    }
    
    .pub-affiliations {
      font-size: 0.9em;
      color: #4b5563;
      margin-bottom: 0.15rem;
    }
    
    .pub-affiliations sup {
      font-size: 0.75em;
      margin-right: 1px;
      color: var(--pub-sup-color);
    }
    
    .pub-venue {
      font-size: 0.9em;
      color: #161616;
      font-weight: 600;
      letter-spacing: 0.01em;
      margin-bottom: 0.25rem;
    }
    
    .pub-links {
      font-size: 0.9em;
      color: #111111;
    }
    
    .pub-links a {
      color: var(--pub-link-color);
      font-weight: 600;
      text-decoration: underline;
      text-decoration-color: var(--pub-link-underline);
      text-underline-offset: 0.14em;
    }

    .pub-links a:hover {
      color: #000000;
      text-decoration-color: var(--pub-link-color);
    }
    
    .people-section {
      display: grid;
      gap: 2rem;
    }

    .people-group h3 {
      font-size: 1rem;
      font-weight: 700;
      color: #1f2937;
      margin-bottom: 0.75rem;
    }

    .people-list {
      list-style: none;
      padding: 0;
      display: grid;
      gap: 0.5rem;
    }

    .people-list li {
      color: #374151;
      font-size: 1rem;
    }

    .people-list a {
      color: #2563eb;
    }

    .people-list a:hover {
      color: #1d4ed8;
    }

    .people-cards {
      list-style: none;
      padding: 0;
      display: flex;
      flex-wrap: wrap;
      gap: 1.25rem 1.25rem;
      margin-bottom: 1.5rem;
    }

    .people-cards li {
      width: 160px;
      text-align: center;
      color: #374151;
      font-size: 0.95rem;
      line-height: 1.35;
      white-space: nowrap;
    }

    .people-cards.text-only {
      gap: 1.25rem 1.25rem;
    }

    .people-cards.text-only li {
      width: auto;
      min-width: 160px;
    }

    .people-cards .name-zh {
      color: #6b7280;
      font-size: 0.9rem;
    }

    .people-cards img {
      width: 120px;
      height: 150px;
      object-fit: contain;
      object-position: center bottom;
      border-radius: 6px;
      display: block;
      margin: 0 auto 0.5rem;
    }

    .people-cards a {
      color: #2563eb;
    }

    .people-cards a:hover {
      color: #1d4ed8;
    }
    
    .group-photos {
      display: flex;
      align-items: flex-start;
      gap: 1.5rem;
      max-width: 760px;
      margin: 0 auto 2rem;
    }

    .group-photo-tabs {
      flex: 0 0 110px;
      display: flex;
      flex-direction: column;
      gap: 0.4rem;
      padding-top: 0.25rem;
      font-size: 0.9rem;
    }

    .group-photo-tabs a {
      color: #6b7280;
      text-decoration: none;
      padding: 0.2rem 0 0.2rem 0.75rem;
      border-left: 2px solid #e5e7eb;
    }

    .group-photo-tabs a:hover {
      color: #1f2937;
    }

    .group-photo-tabs a.active {
      color: #1f2937;
      font-weight: 600;
      border-left-color: #1f2937;
    }

    .group-photos figure {
      flex: 1 1 auto;
      min-width: 0;
      margin: 0;
      display: none;
    }

    .group-photos figure.active {
      display: block;
    }

    .group-photos img {
      width: 100%;
      display: block;
      border-radius: 4px;
    }

    @media (max-width: 640px) {
      .group-photos {
        flex-direction: column;
      }
      .group-photo-tabs {
        flex: none;
        flex-direction: row;
        flex-wrap: wrap;
        gap: 0.75rem;
      }
      .group-photo-tabs a {
        padding: 0 0 2px;
        border-left: none;
        border-bottom: 2px solid transparent;
      }
      .group-photo-tabs a.active {
        border-bottom-color: #1f2937;
      }
    }

    /* Footer */
    footer {
      padding-top: 2rem;
      border-top: 1px solid #e5e7eb;
      font-size: 0.9rem;
      color: #6b7280;
    }
  </style>
</head>
<body>
  <div class="container">
    <!-- Header -->
    <div class="header">
      <h1>Embodied AI (EAI) Lab</h1>
      <div class="header-line"></div>
    </div>
    
    <div class="group-photos">
      <nav class="group-photo-tabs"></nav>
      <figure class="active" data-label="2026 Spring">
        <img src="assets/group_photo/group_photo_2026_spring.jpg" alt="EAI Lab group photo, 2026 Spring">
      </figure>
    </div>
    <script>
      (function () {
        const figures = Array.from(document.querySelectorAll('.group-photos figure'));
        const nav = document.querySelector('.group-photo-tabs');
        figures.forEach((fig, i) => {
          const a = document.createElement('a');
          a.href = '#';
          a.textContent = fig.dataset.label;
          if (fig.classList.contains('active')) a.classList.add('active');
          a.addEventListener('click', (e) => {
            e.preventDefault();
            figures.forEach((f) => f.classList.remove('active'));
            nav.querySelectorAll('a').forEach((t) => t.classList.remove('active'));
            fig.classList.add('active');
            a.classList.add('active');
          });
          nav.appendChild(a);
        });
      })();
    </script>
    
    <!-- Description -->
    <div class="description">
      <p>
        Welcome! The Embodied AI lab are dedicated to developing autonomous, real-world robotic systems. We approach this challenge from three perspectives.  The first is Action Control.  We train robot policies that allow robots to perform mobile manipulation tasks with high efficiency, effectiveness, and robustness.  The second is Learning and Exploration.  We develop algorithms that enable physical agents to explore their environments and continually acquire new skills with minimal human intervention.  The third is Dynamic Modeling.  We build dynamic systems that enhance environmental understanding by simulating the outcomes of physical interactions.  Together, these three research topics aim to create robots that function as intelligently as animal beings.
      </p>
      <p>
        Code for our projects is available at <a href="https://github.com/EmbodiedAI-NTU">https://github.com/EmbodiedAI-NTU</a>.
      </p>
      <p>
        We are looking for Ph.D/Master/undergraduate students interested in robot learning, computer vision, and machine learning. If you are an undergraduate at NTU interested in research positions, please fill out <a href="https://forms.gle/W7rSPF9cM3C5jaZV8">this form</a>. If you are interested in Master/Doctoral positions, please fill out <a href="https://forms.gle/JxzbmXpEAPqp6jkf8">this form</a>.
      </p>

    </div>
    
    <!-- Recent Publications -->
    <section class="publications">
      <h2>Recent Publications <span class="section-note">(<sup>&lowast;</sup> Equal contribution. <sup>&dagger;</sup> Corresponding author.)</span></h2>
      <div class="section-line"></div>
      <ul>
        <li>
          <span class="pub-title">Learning Conditional Source Distribution via Flow Reversal for Temporal Flow Matching</span>
          <div class="pub-authors">
            <strong>Kuan-Hsun Tu</strong><sup>*</sup>, Hsuan-Chi Liu<sup>&dagger;*</sup>, Jia-Wei Liao<sup>*</sup>, Chien-Sheng Chiang, Cheng-Fu Chou, and Tsung-Wei Ke
          </div>
          <div class="pub-venue">ArXiv 2026</div>
          <div class="pub-links"><a href="https://arxiv.org/abs/2610.05349">arXiv</a> / Code / <a href="https://embodiedai-ntu.github.io/cnpflow/index.html">Project Page</a></div>
        </li>
        <li>
          <span class="pub-title">Task Inference Beyond Least Squares in Behavioral Foundation Models</span>
          <div class="pub-authors">
            <strong>Kuan-Hsun Tu</strong><sup>1*</sup>, Chien-Sheng Chiang<sup>1&dagger;*</sup>, Hsin-Wei Chen<sup>1</sup>, Ping-Chun Hsieh<sup>2</sup>, and Tsung-Wei Ke<sup>1</sup>
          </div>
          <div class="pub-affiliations">
            <sup>1</sup>National Taiwan University, <sup>2</sup>National Yang Ming Chiao Tung University
          </div>
          <div class="pub-venue">ArXiv 2026</div>
          <div class="pub-links"><a href="https://arxiv.org/abs/2610.05350">arXiv</a> / Code / <a href="https://embodiedai-ntu.github.io/BLS/index.html">Project Page</a></div>
        </li>
        <li>
          <span class="pub-title">TaRL: Learning General and Physical Rewards from Tactile Demonstrations</span>
          <div class="pub-authors">
            Po-Yi Wu<sup>1</sup>, Dao-Jan Chang<sup>1</sup>, Shang-Ya Hsiao<sup>2</sup>, Hong-Ming Chen<sup>2</sup>, Yu-Cheng Su<sup>2</sup>, Tsung-Wei Ke<sup>1</sup>
          </div>
          <div class="pub-affiliations">
            <sup>1</sup>National Taiwan University, <sup>2</sup>Delta Electronics
          </div>
          <div class="pub-venue">ArXiv 2026</div>
          <div class="pub-links">
            <a href="https://arxiv.org/abs/2609.36785">arXiv</a> / Code / <a href="https://embodiedai-ntu.github.io/tarl/index.html">Project Page</a>
          </div>
        </li>
        <li>
          <span class="pub-title">COT-FM: Cluster-wise Optimal Transport Flow Matching</span>
          <div class="pub-authors">
            Chien-Sheng Chiang<sup>*</sup>, <strong>Kuan-Hsun Tu</strong><sup>*&dagger;</sup>, Jia-Wei Liao<sup>*</sup>, Cheng-Fu Chou, and Tsung-Wei Ke
          </div>
          <div class="pub-venue">CVPR 2026</div>
          <div class="pub-links">
            <a href="https://arxiv.org/abs/2603.13395v1">arXiv</a> / <a href="https://github.com/EmbodiedAI-NTU/COT-FM">Code</a> / <a href="https://embodiedai-ntu.github.io/cotfm/index.html">Project Page</a>
          </div>
        </li>
        <!-- <li>
          <span class="pub-title">CRONOS: Benchmarking Multi-Task Robotic Manipulation for Reset-Free Reinforcement Learning At Scale</span>
          <div class="pub-authors">
            Po Yi Wu<sup>*</sup>, Djengo Cyun-Jyun Fang<sup>*</sup>, Dian Cheng<sup>*</sup>, Tsung-Wei Ke
          </div>
          <div class="pub-venue">In submission</div>
          <div class="pub-links">arXiv / Code / <a href="https://embodiedai-ntu.github.io/cronos/index.html">Project Page</a></div>
        </li> -->
        <li>
          <span class="pub-title">DexMan: Learning Bimanual Dexterous Manipulation from Human and Generated Videos</span>
          <div class="pub-authors">
            Jhen Hsieh<sup>1</sup>, Kuan-Hsun Tu<sup>1</sup>, Kuo-Han Hung<sup>1,2</sup>, Tsung-Wei Ke<sup>1</sup>
          </div>
          <div class="pub-affiliations">
            <sup>1</sup>National Taiwan University, <sup>2</sup>Stanford University
          </div>
          <div class="pub-venue">ArXiv 2025</div>
          <div class="pub-links">
            <a href="https://arxiv.org/abs/2510.08475">arXiv</a> / <a href="https://github.com/EmbodiedAI-NTU/DexMan">Code</a> / <a href="https://embodiedai-ntu.github.io/dexman/index.html">Project Page</a>
          </div>
        </li>
        <li>
          <span class="pub-title">Information Seeking for Robust Decision Making under Partial Observability</span>
          <div class="pub-authors">
            Djengo Cyun-Jyun Fang<sup>1</sup>, Tsung-Wei Ke<sup>1</sup>
          </div>
          <div class="pub-affiliations">
            <sup>1</sup>National Taiwan University
          </div>
          <div class="pub-venue">ArXiv 2025</div>
          <div class="pub-links">
            <a href="https://arxiv.org/abs/2510.01531">arXiv</a> / <a href="https://github.com/InfoSeekerLLM/InfoSeeker">Code</a> / <a href="https://infoseekerllm.github.io/">Project Page</a>
          </div>
        </li>
      </ul>
    </section>
    
    <section>
      <h2>People</h2>
      <div class="section-line"></div>
      <div class="people-section">
        <div class="people-group">
          <h3>Professor</h3>
          <ul class="people-cards">
            <li><a href="https://twke18.github.io/"><img src="assets/people/tsung_wei_ke.jpg" alt="Tsung-Wei Ke">Tsung-Wei Ke<br><span class="name-zh">(柯宗瑋)</span></a></li>
          </ul>
        </div>
        <div class="people-group">
          <h3>Master Students</h3>
          <ul class="people-cards">
            <li><a href="https://alexpeng517.github.io/"><img src="assets/people/li_xuan_peng.jpg" alt="Li-Xuan Peng">Li-Xuan Peng<br><span class="name-zh">(彭立軒)</span></a></li>
            <li><img src="assets/people/ruei_jia_huang.jpg" alt="Ruei-Jia Huang">Ruei-Jia Huang<br><span class="name-zh">(黃睿加)</span></li>
          </ul>
          <ul class="people-cards text-only">
            <li>Fang Jeff<br><span class="name-zh">(方世昕)</span></li>
            <li>Dao-Jan Chang<br><span class="name-zh">(張道然)</span></li>
          </ul>
        </div>
        <div class="people-group">
          <h3>Undergraduate Students</h3>
          <ul class="people-cards">
            <li><a href="https://syun413.github.io/"><img src="assets/people/kuan_hsun_tu.jpg" alt="Kuan-Hsun Tu">Kuan-Hsun Tu<br><span class="name-zh">(杜冠勳)</span></a></li>
            <li><img src="assets/people/chien_sheng_chiang.jpg" alt="Chien-Sheng Chiang">Chien-Sheng Chiang<br><span class="name-zh">(江建陞)</span></li>
            <li><img src="assets/people/jhen_hsieh.jpg" alt="Jhen Hsieh">Jhen Hsieh<br><span class="name-zh">(謝震)</span></li>
            <li><img src="assets/people/chun_yang_cheng.jpg" alt="Chun-Yang Cheng">Chun-Yang Cheng<br><span class="name-zh">(鄭峻揚)</span></li>
            <li><img src="assets/people/wei_lun_hsiao.jpg" alt="Wei-Lun Hsiao">Wei-Lun Hsiao<br><span class="name-zh">(蕭葦倫)</span></li>
            <li><img src="assets/people/daisy_yuhsin_huang.jpg" alt="Daisy Yuhsin Huang">Daisy Yuhsin Huang<br><span class="name-zh">(黃予星)</span></li>
            <li><img src="assets/people/hsuan_chi_liu.jpg" alt="Hsuan-Chi Liu">Hsuan-Chi Liu<br><span class="name-zh">(劉軒齊)</span></li>
            <li><img src="assets/people/li_chen_kao.jpg" alt="Li-Chen Kao">Li-Chen Kao<br><span class="name-zh">(高立辰)</span></li>
          </ul>
          <ul class="people-cards text-only">
            <li>Yen-Hsiang Chen<br><span class="name-zh">(陳彥翔)</span></li>
            <li>Dian Cheng<br><span class="name-zh">(鄭琠)</span></li>
            <li>Hsin-Wei Chen<br><span class="name-zh">(陳忻韋)</span></li>
            <li>Yu-Tsen Tsai<br><span class="name-zh">(蔡侑岑)</span></li>
            <li>We-Hsiang Chiu<br><span class="name-zh">(邱威翔)</span></li>
            <li>Po-Ke Fan<br><span class="name-zh">(范柏格)</span></li>
            <li>Cheng-Ting Tsai<br><span class="name-zh">(蔡政廷)</span></li>
          </ul>
        </div>
        <div class="people-group">
          <h3>Alumni</h3>
          <ul class="people-cards text-only">
            <li>Pang-Chi Lo<br><span class="name-zh">(羅邦齊)</span></li>
            <li>Jhao-Hsun Lai<br><span class="name-zh">(賴昭勳)</span></li>
            <li>Djengo Cyun-Jyun Fang<br><span class="name-zh">(方群鈞)</span></li>
            <li>Po-Yi Wu<br><span class="name-zh">(吳柏毅)</span></li>
          </ul>
        </div>
      </div>
    </section>
    
    <!-- Footer -->
    <footer>
      © 2026 Embodied AI Lab · National Taiwan University
    </footer>
  </div>
</body>
</html>