Xiangyu Chen

MARS Lab, School of MAE

Nanyang Technological University

Singapore

I am a Ph.D. student in the Multimodal AI Research Society (MARS) Lab, School of Mechanical and Aerospace Engineering, Nanyang Technological University (NTU), Singapore, advised by Prof. Jianfei Yang (Stanford World’s Top 2% Scientists). I also collaborate closely with Prof. Chen Wang and Prof. Chaoqun Wang.

My research centers on robotic manipulation and dexterous manipulation. I am currently building embodied AI agents for life science that pair foundation models with robotic arms to automate Chem/Bio laboratory work. My goal is a robust, safe, and easy-to-use agent that takes over the repetitive bench tasks for biochemical researchers — much like Copilot assists programmers and ChatGPT assists writers.

Before joining NTU, I earned my M.Eng. in Control Science and Engineering from Shandong University and my B.Eng. in Electrical Engineering and Automation from the University of Jinan, where I worked on practical applications of mobile robots.

Outside of research, I am a keen basketball player and enjoy video games — especially Roguelikes and Pokémon — and, like many Ph.D. students, I have recently taken up photography.

Feel free to email me if our interests align — I am always happy to discuss research and potential collaborations.

news

Apr 15, 2026	🎉 Our paper “A2A: Action-to-Action Flow Matching for Robot Policy Learning” is accepted by RSS 2026!
Feb 27, 2026	🎉 Our paper “M³A Policy: Mutable Material Manipulation Augmentation Policy through Photometric Re-rendering” is accepted by CVPR 2026 (Findings)!
Jan 15, 2026	🎉 Our paper “RM-RL: Role-Model Reinforcement Learning for Precise Robot Manipulation” is accepted by ICRA 2026!
Oct 01, 2025	🎉 Our paper *“iA: Imperative Learning-based A* Search for Pathfinding” is accepted by IEEE RA-L**!
Dec 15, 2024	🎓 Received the Ph.D. offer from NTU — beginning my Ph.D. study in 2025!

selected publications

arXiv

Rethinking Implicit Spatial Representation in Visuomotor Policy Learning

Xiangyu Chen, Yuxuan Hu, Chuhao Zhou, and Jianfei Yang

arXiv preprint arXiv:2606.15232, Jun 2026

arXiv Bib PDF

@article{chen2026rethinking,
  title = {Rethinking Implicit Spatial Representation in Visuomotor Policy Learning},
  author = {Chen, Xiangyu and Hu, Yuxuan and Zhou, Chuhao and Yang, Jianfei},
  journal = {arXiv preprint arXiv:2606.15232},
  year = {2026},
  month = jun,
}

ICRA

RM-RL: Role-Model Reinforcement Learning for Precise Robot Manipulation

Xiangyu Chen, Chuhao Zhou, Yuxi Liu, and Jianfei Yang

In IEEE International Conference on Robotics and Automation (ICRA), Jan 2026

arXiv Bib PDF Website

@inproceedings{chen2026rmrl,
  title = {RM-RL: Role-Model Reinforcement Learning for Precise Robot Manipulation},
  author = {Chen, Xiangyu and Zhou, Chuhao and Liu, Yuxi and Yang, Jianfei},
  booktitle = {IEEE International Conference on Robotics and Automation (ICRA)},
  year = {2026},
  month = jan,
}

RA-L

iA*: Imperative Learning-based A* Search for Pathfinding

Xiangyu Chen, Fan Yang, and Chen Wang

IEEE Robotics and Automation Letters (RA-L), Oct 2025

arXiv Bib PDF Code Website

@article{chen2025iastar,
  title = {iA*: Imperative Learning-based A* Search for Pathfinding},
  author = {Chen, Xiangyu and Yang, Fan and Wang, Chen},
  journal = {IEEE Robotics and Automation Letters (RA-L)},
  year = {2025},
  month = oct,
}