index.html

<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<meta name="theme-color" content="#f8f8f5">
<title>Muyang Chen | Multimodal AI Research</title>
<meta name="description" content="Muyang Chen is an undergraduate researcher at CUHK-Shenzhen working on multimodal reasoning, intent recognition, and affective computing.">
<link rel="canonical" href="https://sv-arcode.github.io/">
<style>
:root{--ink:#232c32;--muted:#59656c;--accent:#245e57;--line:#dce2de;--paper:#f8f8f5;--soft:#edf2ee}
*{box-sizing:border-box}
html{scroll-behavior:smooth;scroll-padding-top:28px;-webkit-text-size-adjust:100%}
body{margin:0;background:var(--paper);color:var(--ink);font:16px/1.7 "Segoe UI",-apple-system,BlinkMacSystemFont,"Helvetica Neue",Arial,sans-serif}
a{color:var(--accent);text-underline-offset:4px;text-decoration-thickness:1px}
a:hover{text-decoration-thickness:2px}
a:focus-visible,summary:focus-visible{outline:3px solid var(--accent);outline-offset:5px;border-radius:2px}
.skip{position:absolute;left:20px;top:-100px;padding:10px 16px;background:#fff;z-index:10}
.skip:focus{top:12px}
.layout{max-width:1160px;margin:auto;padding:64px 40px 40px;display:grid;grid-template-columns:240px minmax(0,1fr);gap:76px}
.identity{position:sticky;top:44px;align-self:start}
.eyebrow{font-size:11px;letter-spacing:.13em;text-transform:uppercase;color:var(--muted);font-weight:600;margin:0 0 16px}
h1{font-family:Georgia,"Times New Roman",serif;font-weight:400;font-size:43px;line-height:1.1;letter-spacing:-1.5px;margin:0 0 10px}
.cn{font-size:18px;letter-spacing:.12em;color:var(--muted);margin:0 0 18px}
.affiliation{font-size:14px;line-height:1.65;color:var(--muted);margin:0 0 24px}
.contacts{display:flex;gap:10px;flex-wrap:wrap;margin-bottom:24px}
.contacts a{display:inline-flex;align-items:center;min-height:44px;padding:7px 12px;border:1px solid var(--line);border-radius:4px;text-decoration:none;font-size:13px;font-weight:600;background:#fff}
.contacts a.primary{background:var(--accent);color:#fff;border-color:var(--accent)}
.contacts a:hover{border-color:var(--accent)}
.email{font-size:13px;overflow-wrap:anywhere}
.section-nav{border-top:1px solid var(--line);padding-top:22px;margin-top:28px;display:flex;flex-direction:column;gap:3px}
.section-nav a{font-size:14px;padding:7px 0;text-decoration:none;color:var(--muted)}
.section-nav a:hover{color:var(--accent)}
main{min-width:0}
section{margin-bottom:48px;scroll-margin-top:24px}
h2{font-family:Georgia,"Times New Roman",serif;font-size:27px;line-height:1.3;font-weight:400;letter-spacing:-.4px;margin:0 0 22px}
h3{font-size:18px;line-height:1.45;letter-spacing:-.2px;margin:0 0 5px;font-weight:650}
p{margin:0 0 15px}
.intro{font-size:18px;line-height:1.75}
.intro strong{font-weight:600}
.opportunity{border-left:3px solid var(--accent);padding:10px 16px;background:var(--soft);font-size:14px;margin:24px 0 0}
.interests{font-size:13px;color:var(--muted);margin-top:20px}
.interests span{display:inline-block;margin-right:14px}
.section-title{display:flex;align-items:baseline;justify-content:space-between;gap:16px;border-bottom:1px solid var(--line);margin-bottom:24px;padding-bottom:12px}
.section-title h2{margin:0}
.section-title a{font-size:13px;white-space:nowrap}
.project{padding:0 0 28px;margin-bottom:28px;border-bottom:1px solid var(--line)}
.project:last-child{margin-bottom:0;padding-bottom:0;border:0}
.meta{font-size:13px;color:var(--muted);margin:0 0 14px}
.status{font-size:12px;font-weight:600;color:var(--accent);margin:0 0 8px;letter-spacing:.02em}
.project p,.project li{font-size:15px}
.project .meta{font-size:13px}
.project .status{font-size:12px}
ul{padding-left:20px;margin:12px 0 0}
li{padding-left:3px;margin-bottom:7px}
.result{color:var(--accent);font-size:14px!important;margin-top:14px}
details{margin-top:16px;font-size:14px}
summary{color:var(--accent);cursor:pointer;padding:7px 0;width:fit-content;font-size:13px}
details p{color:var(--muted);font-size:14px!important;margin-top:10px}
.education-head{margin-bottom:18px}
.grades{display:grid;grid-template-columns:repeat(3,minmax(0,1fr));gap:16px;margin:20px 0 24px;padding:20px 0;border-top:1px solid var(--line);border-bottom:1px solid var(--line)}
.grades dt{font-size:12px;color:var(--muted);line-height:1.5;margin-bottom:6px}
.grades dd{margin:0;font-size:22px;font-family:Georgia,"Times New Roman",serif}
.grades dd span{font-size:13px;font-family:"Segoe UI",sans-serif;color:var(--muted)}
.small{font-size:14px}
.honors{font-size:14px;margin-top:14px}
.skills{display:grid;grid-template-columns:110px minmax(0,1fr);gap:8px 18px;font-size:14px;margin:0}
.skills dt{color:var(--muted)}
.skills dd{margin:0}
footer{border-top:1px solid var(--line);padding-top:20px;display:flex;justify-content:space-between;gap:15px;flex-wrap:wrap;color:var(--muted);font-size:12px}
@media(max-width:850px){.layout{grid-template-columns:200px minmax(0,1fr);gap:36px;padding:40px 28px}.identity h1{font-size:36px}.contacts{gap:7px}.contacts a{padding:7px 10px}}
@media(max-width:650px){.layout{display:block;padding:32px 22px}.identity{position:static;margin-bottom:32px}.identity h1{font-size:42px}.cn{display:inline;font-size:17px}.affiliation{margin:14px 0 18px}.contacts{margin-bottom:10px}.section-nav{flex-direction:row;flex-wrap:wrap;gap:0 22px;margin-top:20px;padding-top:10px}.section-nav a{min-height:44px}.eyebrow{margin-bottom:12px}.intro{font-size:17px}section{margin-bottom:38px}.grades{gap:10px}.grades dd{font-size:21px}.section-title{gap:10px}.section-title h2{font-size:25px}.skills{grid-template-columns:90px minmax(0,1fr);gap:8px 12px}}
@media(prefers-reduced-motion:reduce){html{scroll-behavior:auto}}
@media print{body{background:#fff;font-size:11pt}.layout{display:block;padding:0;max-width:none}.identity{position:static;margin-bottom:24px}.section-nav,.skip,details{display:none}.contacts{margin:12px 0}.project{break-inside:avoid}section{margin-bottom:24px}a{color:inherit}.grades{padding:10px 0}footer{margin-top:20px}}
</style>
</head>
<body>
<a class="skip" href="#main">Skip to content</a>
<div class="layout">
<header class="identity">
<p class="eyebrow">AI · Multimodal reasoning</p>
<h1>Muyang Chen</h1>
<p class="cn" lang="zh-CN">陈沐阳</p>
<p class="affiliation">Undergraduate researcher<br>The Chinese University of Hong Kong, Shenzhen</p>
<div class="contacts" aria-label="Contact and résumé">
<a class="primary" href="CV_MuyangChen.pdf">CV (PDF)</a>
<a href="mailto:muyangchen@link.cuhk.edu.cn">Email</a>
<a href="https://github.com/Sv-Arcode">GitHub</a>
</div>
<a class="email" href="mailto:muyangchen@link.cuhk.edu.cn">muyangchen@link.cuhk.edu.cn</a>
<nav class="section-nav" aria-label="Page sections">
<a href="#about">About</a>
<a href="#research">Research &amp; projects</a>
<a href="#education">Education</a>
<a href="#teaching">Teaching</a>
<a href="#skills">Skills</a>
</nav>
</header>
<main id="main">
<section id="about" aria-labelledby="about-title">
<h2 id="about-title">About</h2>
<p class="intro">I study how multimodal models use evidence to understand intent and emotion. My interests are in <strong>multimodal reasoning, intent recognition, and affective computing</strong>.</p>
<p>I am a final-year Electrical and Computer Engineering student at CUHK-Shenzhen and an undergraduate researcher in the NAIS Lab. My lab PI is Prof. Fangxin Wang, and my research mentor is Dr. Dayou Zhang.</p>
<p>Through my work on MISID and AffectiveArt, I have worked on annotation, model evaluation, and multimodal systems. I am interested in understanding where models fail and how to evaluate their reasoning more carefully.</p>
<p>I expect to graduate in June 2027.</p>
</section>
<section id="research" aria-labelledby="research-title">
<div class="section-title"><h2 id="research-title">Research &amp; projects</h2><a href="CV_MuyangChen.pdf">Full CV ↗</a></div>
<article class="project">
<p class="status">Co-first author · Under review at AAAI 2027</p>
<h3>MISID: Multimodal Intent Recognition</h3>
<p class="meta">NAIS Lab, CUHK-Shenzhen · Sep 2025 – Present</p>
<p>A dataset and framework for studying hidden intent in long, multimodal conversations in strategic deception games.</p>
<ul>
<li>Co-developed the annotation scheme for 3,962 aligned audiovisual utterance segments and independently evaluated inter-annotator agreement.</li>
<li>Designed FRACTAM's modal decoupling and text reconstruction component for annotation and inference.</li>
<li>Contributed to model evaluation, conducted ablation studies, and co-designed the LLM-as-a-Judge protocol and FRACTAM framework.</li>
</ul>
<p class="result">The full framework improved mean Hidden Intent Inference Score from 34.22 to 42.79 across ten language-model backbones.</p>
<details><summary>Method and evaluation details</summary>
<p>The modal decoupling component pairs audio with eGeMAPS features and sampled video frames with FACS descriptors. It generates modality-specific descriptions without additional speech transcripts to limit textual interference and support subsequent cross-modal inconsistency analysis.</p>
<p>Inter-annotator agreement reached Krippendorff's α of 0.80–0.94 across structured fields. I contributed to the zero-shot evaluation of 19 language and video-language models. FRACTAM combines modality decoupling, long-range retrieval, and cross-modal evidence chains.</p>
</details>
</article>
<article class="project">
<p class="status">Team lead · 5th of 14 teams</p>
<h3>AffectiveArt: Artwork Emotion Understanding</h3>
<p class="meta">ACM Multimedia 2026 Grand Challenge, Track 2 · May – Jun 2026</p>
<p>A system for recognizing emotion in artworks and generating emotion descriptions.</p>
<ul>
<li>Led a three-member team and designed the overall system architecture.</li>
<li>Fine-tuned image classifiers and designed validation-weighted voting, which outperformed the best individual model.</li>
<li>Independently fine-tuned SigLIP2 and DINOv3 for visual-attribute salience prediction; salience and classification outputs conditioned a LoRA-fine-tuned Qwen3.5-27B description model.</li>
</ul>
<details><summary>Models and results</summary>
<p>The classifiers used Gemma, Qwen, SigLIP2, and DINOv3. The system achieved 94.0% arousal, 88.3% valence, and 79.2% emotion accuracy on the hidden test set. Generated descriptions are outputs, not inputs to the classifiers.</p>
</details>
</article>
<article class="project">
<p class="status">Course project · Project lead</p>
<h3>Interpretable Butterfly Classification</h3>
<p class="meta">ECE4513 Image Processing and Computer Vision · Summer 2025</p>
<p>Led a four-layer CNN project to classify ten visually similar butterfly species, achieving 83.23% test accuracy. Implemented Class Activation Mapping from scratch and used PCA of feature maps to examine the image regions supporting predictions.</p>
<a class="small" href="https://github.com/Sv-Arcode/butterfly-cnn-interpretability">View code on GitHub ↗</a>
</article>
</section>
<section id="education" aria-labelledby="education-title">
<div class="section-title"><h2 id="education-title">Education</h2></div>
<div class="education-head"><h3>The Chinese University of Hong Kong, Shenzhen</h3>
<p class="small">B.Eng. in Electrical and Computer Engineering<br>Computer Engineering stream</p>
<p class="meta">Sep 2023 – Jun 2027 (expected)</p></div>
<dl class="grades">
<div><dt>Cumulative GPA</dt><dd>3.81 <span>/ 4.00</span></dd></div>
<div><dt>Major GPA</dt><dd>3.90 <span>/ 4.00</span></dd></div>
<div><dt>Major GPA rank · CE stream</dt><dd>3 / 60 <span>· Top 5%</span></dd></div>
</dl>
<p class="small"><strong>Selected coursework (all A):</strong> Machine Learning; Machine Intelligence and Applications; Image Processing and Computer Vision; Parallel Programming.</p>
<ul class="honors">
<li>Dean's List, School of Science and Engineering: 2023–24, 2024–25, 2025–26.</li>
<li>Academic Specialty Entrance Scholarship, Second Class: 2023 (renewable).</li>
</ul>
</section>
<section id="teaching" aria-labelledby="teaching-title">
<div class="section-title"><h2 id="teaching-title">Teaching</h2></div>
<h3>Undergraduate Student Teaching Fellow</h3>
<p class="meta">MAT1001 Calculus I, CUHK-Shenzhen · Sep – Dec 2024</p>
<p class="small">Supported weekly tutorials and office hours for approximately 60 students.</p>
</section>
<section id="skills" aria-labelledby="skills-title">
<div class="section-title"><h2 id="skills-title">Skills &amp; languages</h2></div>
<dl class="skills">
<dt>Programming</dt><dd>Python, C++</dd>
<dt>ML &amp; vision</dt><dd>PyTorch, OpenCV</dd>
<dt>Tools</dt><dd>Git, Linux, LaTeX</dd>
<dt>Languages</dt><dd>Mandarin (native); English (IELTS Academic 7.0)</dd>
</dl>
</section>
<footer><span>© 2026 Muyang Chen</span><span>Updated 19 September 2026 · <a href="#about">Back to top ↑</a></span></footer>
</main>
</div>
</body>
</html>