Hello! I am currently a PhD student in Zhejiang University (浙江大学), majoring in Computer Science and Technology supervised by Prof. Jianke Zhu. I have obtained B.Eng. (with Honors) from Wuhan University (武汉大学) majoring in Computer Science and Technology supervised by Prof. Zheng Wang . I used to be a summer research intern at McGill University and Mila-Quebec AI Institute in Montreal, Canada, under the supervision of Prof. Xujie Si. Prior to that, I was a visit student at KAIST in Daejeon, Korea, supervised by Prof. Chang D. Yoo.

My research interests include 2D/3D Multimodal LLMs, Visual/Scene Understanding, Spatial Intelligence, and Embodied AI, particularly in:

1.Enabling MLLMs with common visual tasks, including open-vocabulary visual grounding for image/video/3D scene.

2.Embodied scene understanding/reasoning, including streaming 3D interaction and embodied dialogue/planning.

3.Efficient and effective MLLMs, including visual token compression and lightweight MLLM.

If you are interested in any form of academic cooperation with me, please feel free to email at hanxun.yu@zju.edu.cn.

🔥 News

2026.01: 🎉🎉 One paper is accepted by ICLR 2026.
2025.02: 🎉🎉 One paper is accepted by CVPR 2025 Highlight. (2.9%, 387/13008)
2024.07: 🎉🎉 One paper is accepted by IEEE TPAMI 2024.
2023.07: 🎉🎉 One paper is accepted by ACM MM 2023.
2023.06: 🎉🎉 I won the National Scholarship at Wuhan University. (Top 2%)
2022.06: 🎉🎉 Accepted to the Mitacs Globalink Research Internship 2022 program. (200/year Nationwide)

📝 Selected Publications

* indicates equal contribution

ICLR 2026

Hanxun Yu*, Wentong Li*, Xuan Qu*, Song Wang, Junbo Chen, Jianke Zhu

ICLR 2026

An efficient vision token compression framework with two modules, Dominant Vision Token Selection (DVTS) and Text-Guided Vision Complement (TGVC).

CVPR 2025 (Highlight)

Hanxun Yu*, Wentong Li*, Song Wang, Junbo Chen, Jianke Zhu

CVPR 2025 (Highlight, Top 2.9%)

A unified and effective instance-aware 3D Large Multi-modal Model for multi-task 3D scene understanding through coupled 2D-3D modality encoding.

ACM MM 2023

Hui Wei*, Hanxun Yu*, Kewei Zhang, Zhixiang Wang, Jianke Zhu, Zheng Wang

ACM MM 2023

This paper focuses on AI safety-critical tasks and innovatively proposes moiré-based backdoor attack triggers into pedestrian detection models.

arXiv

Yuxin Wang, Lei Ke, Boqiang Zhang, Tianyuan Qu, Hanxun Yu, Zhenpeng Huang, Meng Yu, Dan Xu, Dong Yu

arXiv 2025

[Project] [Paper] [Code]

A unified framework that empowers native 3D grounding to enable accurate spatial reasoning in Vision-Language Models.

arXiv

Xinqi Jin*, Hanxun Yu*, Bohan Yu, Kebin Liu, Jian Liu, Keda Tao, Yixuan Pei, Huan Wang, Fan Dang, Jiangchuan Liu, Weiqiang Wang

arXiv 2025

[Paper] [Code]

A token pruning method designed to reduce both spatial and temporal redundancy in online video understanding.

TPAMI 2024

Hui Wei, Hao Tang, Xuemei Jia, Zhixiang Wang, Hanxun Yu, Zhubo Li, Shin’ichi Satoh, Luc Van Gool, Zheng Wang

IEEE TPAMI 2024

This survey aims to summarize existing physical adversarial attack methods, providing insights toward the development of trustworthy AI systems.

2024 The Chiang Chen Scholarship, China.
2024,2025 The First Prize of Excellent Graduate Scholarship, Zhejiang University.
2023 The National Scholarship, China. (Top 2%)
2023 Outstanding Undergraduate Dissertation Award, Wuhan University.
2023 Outstanding Graduate, Wuhan University.
2022 Mitacs-CSC Globalink Research Internship Scholarship, China. (200/year Nationwide)
2020,2021,2022 The First Prize of Excellent Undergraduate Scholarship, Wuhan University.