Henry Hao-Tang Tsui

henrytsu [at] andrew.cmu.edu

Hi! I’m Hao-Tang Tsui, feel free to call me Henry!

I work on computer vision, vision–language models, and trying to build a usable bridge between pixels and words. I’m currently a Master’s student in Computer Vision at Carnegie Mellon University, working with Prof. Deva Ramanan on vision–language benchmarking and Prof. Kris Kitani on 3D generation.

Previously, I was a research assistant at Academia Sinica (YOLO-Lab) with Prof. Mark Liao, where I worked on YOLO-related research and re-release YOLO under the MIT license. I earned my B.S. in Electrical Engineering from National Yang Ming Chiao Tung University, collaborating with Prof. Hong-Han Shuai and Prof. Wen-Huang Cheng.

news

Sep 29, 2026	My paper Point2Part is now on arXiv!
Jun 05, 2026	My paper Σ StaDy4D received the Best Paper Award at the CVPR 2026 GenRecon3D Workshop!
May 13, 2026	My paper Σ StaDy4D was selected for an Oral at the CVPR 2026 GenRecon3D Workshop!
Apr 05, 2026	Our paper TTSG was accepted by CVPR 2026 Workshop!
Dec 31, 2025	My code YOLO-MIT is now available on GitHub!

selected publications

arXiv

Point2Part: Unified 3D Partitioning from Point Prompts

Hao-Tang Tsui, Yu-Rou Tuan, Xiaoxuan Ma, and 3 more authors

Under review , Sep 2026

arXiv Bib PDF Website

@misc{tsui2026point2part,
  author = {Tsui, Hao-Tang and Tuan, Yu-Rou and Ma, Xiaoxuan and Ugrinovic, Nicolas and Shiratori, Takaaki and Kitani, Kris},
  title = {Point2Part: Unified 3D Partitioning from Point Prompts},
  year = {2026},
  month = sep,
  eprint = {2609.38180},
  archiveprefix = {arXiv},
  primaryclass = {cs.CV},
}

ICLR

YOLO-RD: Introducing Relevant and Compact Explicit Knowledge to YOLO by Retriever-Dictionary

Hao-Tang Tsui, Chien-Yao Wang, and Hong-Yuan Mark Liao

In Proceedings of the International Conference on Learning Representations (ICLR), Apr 2025

arXiv Bib PDF Code Website

@inproceedings{tsui2024yolord,
  author = {Tsui, Hao-Tang and Wang, Chien-Yao and Liao, Hong-Yuan Mark},
  title = {YOLO-RD: Introducing Relevant and Compact Explicit Knowledge to YOLO by Retriever-Dictionary},
  booktitle = {Proceedings of the International Conference on Learning Representations (ICLR)},
  year = {2025},
  month = apr,
  eprint = {2410.15346},
  archiveprefix = {arXiv},
  primaryclass = {cs.CV},
}