Ruixiang Wang

Ph.D. Student · CUHK-SZ

[Google Scholar] [GitHub]

About

I am a Ph.D. student in Computer Science at the School of Data Science, The Chinese University of Hong Kong, Shenzhen (CUHK-Shenzhen), advised by Prof. Kui Jia. Previously, I worked with Prof. Guyue Zhou at the Institute for AI Industry Research (AIR), Tsinghua University. I also conducted research at King Abdullah University of Science and Technology (KAUST) under the supervision of Prof. Mohamed Elhoseiny and worked as a Research Intern at DexForce Technology. Before that, I received my B.Eng. in Automation with honors from Harbin Institute of Technology, Weihai.

My research focuses on building embodied intelligence for perceiving and interacting with the physical world. In particular, I am interested in developing generalizable robotic manipulation policies and learning world models that capture how the physical world appears, behaves, and evolves.

Feel free to contact me for discussion or collaboration!

News

  • [2026.10] We release UNITAS, a 3D-Native world action model.
  • [2026.06] Our team won first place in the WorldArena Challenge at CVPR 2026. Champion
  • [2026.06] One paper was accepted to IROS 2026.
  • [2026.04] YOTO++ was accepted to IEEE TPAMI.
  • [2026.03] EVA was released and open-sourced.
  • [2025.10] One paper was accepted to IEEE Robotics and Automation Letters (RA-L).
  • [2025.09] I started my Ph.D. at The Chinese University of Hong Kong, Shenzhen.
  • [2025.06] GAT-Grasp was accepted to IROS 2025.
  • [2025.06] One paper was accepted to ICCV 2025.
  • [2025.04] YOTO was accepted to RSS 2025.

Education

Experience

Publications

UNITAS overview: diverse embodiments, a shared 3D world-coordinate interface, and performance

UNITAS: A 3D-Native World Action Model for Embodied Manipulation

Ruixiang Wang*, Yongyi Su*†, Wenlve Zhou, Bo Yue, Hengyan Liu, Dekun Lu, Yuxin Tian, Yihan Fang, Zerui Wu, Xing Hu, Jietao Chen, Yong Guo, Ziyan He, Junbin Yuan, Guiliang Liu, Xiaofen Xing, Kui Jia

arXiv preprint, 2026

[Paper] [Project Page] [Code]

Vid2WAM framework overview

Vid2WAM: Distilling Video Diffusion Priors into World Action Models

Chenhao Qiu*, Ruixiang Wang*†, Runyi Zhao*, Sixu Lin, Songen Gu, Shufeng Nan, Guiliang Liu, Kui Jia, Yanwei Fu, Simo Wu (* equal contribution · † project lead)

arXiv preprint, 2026

[Paper] [Project Page]

EVA overview

EVA: Aligning Video World Models with Executable Robot Actions via Inverse Dynamics Rewards

Ruixiang Wang, Qingming Liu, Yueci Deng, Guiliang Liu, Zhen Liu, Kui Jia

arXiv preprint, 2026

[Paper] [Project Page] [Code]

YOTO++ system and task overview

YOTO++: Learning Long-Horizon Closed-Loop Bimanual Manipulation from One-Shot Human Video Demonstrations

Huayi Zhou, Ruixiang Wang, Yunxin Tai, Yueci Deng, Guiliang Liu, Kui Jia

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), 2026

[Paper] [Project Page] [Code]

YOTO teaser

You Only Teach Once: Learn One-Shot Bimanual Robotic Manipulation from Video Demonstrations

Huayi Zhou, Ruixiang Wang, Yunxin Tai, Yueci Deng, Guiliang Liu, Kui Jia

Robotics: Science and Systems (RSS), 2025

[Paper] [Project Page] [Code]

GAT-Grasp overview

GAT-Grasp: Gesture-Driven Affordance Transfer for Task-Aware Robotic Grasping

Ruixiang Wang, Huayi Zhou, Xinyue Yao, Guiliang Liu, Kui Jia

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2025

[Paper] [arXiv]

Research Interests

01

Robot Manipulation

Generalizable policies for reliable, long-horizon manipulation.

02

World Models

Predictive models that remain physically grounded and executable.

03

Embodied Intelligence

Connecting perception, imagination, and action in the physical world.