Yao Xiao

CS Ph.D. candidate @ UIUC

Google Scholar
GitHub
X
Linkedin

About Me

I am a final-year Ph.D. candidate in Computer Science at University of Illinois Urbana-Champaign, advised by Prof. Derek Hoiem.

My current research interests mainly lie in computer vision, especially in:

  • Closed-loop Planning: Complex tasks demand sequential decision-making under partial observability, where no single observation reveals the full state, and every decision must build on the history of what came before.

  • Memory Representation: How should an agent encode its past into a compact state, one that helps the present without drowning it, and can be updated if needed?

  • Experience Retrieval: How do we recall exactly the right piece of the past at minimum cost?

I'm always open to academic discussions and new collaborations. If you’re interested in my research or would simply like to connect, feel free to send me an email!

Selected Work

For the full publication list, please refer to my Google Scholar.

ReToken: One Token to Improve Vision–Language Models for Visual Retrieval

Yao Xiao, Reuben Tan, Zhen Zhu, Yuqun Wu, Jianfeng Gao, Derek Hoiem

Under Review

[TL;DR]
Attention tells you where a VLM looks, so shouldn't it tell you which frames matter? It doesn't: the retrieval signal lives in the value space, and a single learned token is enough to read it out.
[Paper] [Code]
ECCV 2026🏅Oral

How to Teach Large Multimodal Models New Skills?

Zhen Zhu, Yiming Gong, Yao Xiao, Yaoyao Liu, Derek Hoiem

ECCV, 2026. Oral.

[Paper] [Code]
TMLR 2025🏅J2C

TextRegion: Text-Aligned Region Tokens from Frozen Image-Text Models

Yao Xiao, Qiqian Fu, Heyi Tao, Yuqun Wu, Zhen Zhu, Derek Hoiem

TMLR, 2025. J2C certification (top 10%).

[TL;DR]
A simple, training-free approach to get region tokens directly comparable to text embeddings, achieving state-of-the-art zero-shot region-level understanding.
[Paper] [Code]

Blog

A casual blog for sharing thoughts and tracking my journey. No tech talk, just real life.

→ Full list

Teaching

Services

Co-organizer: UIUC Computer Vision External Speaker SeriesInterested speakers are welcome to reach out to register for upcoming sessions.

Reviewer: CVPR, ICCV, ECCV, ICLR, NeurIPS, WACV, IEEE Access.

Last Updated: 7/30/2026, 9:04:32 PM