DreaMax
<head>
<meta name="viewport" content="width=device-width, initial-scale=1.0">
<title>What do I do</title>
</head>
<body>
<h1>Things I do to get a perfect background image</h1>
<p>Maybe I should stop tinkering with VSCode settings
just to take a screenshot of this dummy html code.</p>
<span>Oops, Almost forgot to say "Hello World!".</span>
</body>
</html>
Undergraduate Researcher | Graph ML, Algorithms & Agentic World Models
"It’s about a man who has nothing, who risks everything, to feel something." — Jessica Day
A website
for everyone
to learn and
change the world!
< version 1.0 >
A website for everyone to learn and change the world!
< version 1.0 >
Selected Research
Deep Residual Learning for Image Recognition
Proposed residual learning framework that enables training of substantially deeper networks by reformulating layers as learning residual functions with reference to layer inputs.
NeoVerse: Enhancing 4D World Model with in-the-wild Monocular Videos
A novel framework that leverages diverse monocular video data to improve 4D world model generalization across complex dynamic scenes.
MegaSaM: Accurate, Fast and Robust Structure and Motion from Casual Dynamic Videos
A robust system for recovering accurate camera parameters and dense scene structure from casually captured dynamic videos with fast inference speed.
Projects
Project Name
A short description of what this project does.
Another Project
What this project is about in one line.
Recent Updates
Latest Update Title
What was added or updated.
Previous Update
What was added or updated.
Earlier Update
What was added or updated.