Jiashun Wang

I am a final-year PhD student at Robotics Institute, School of Computer Science at Carnegie Mellon University, advised by Prof. Jessica Hodgins. I obtained my master degree at UC, San Diego, advised by Prof. Xiaolong Wang. I received my B.S. degree in Statistics and Data Science from Fudan University. I have interned at NVIDIA, working with Prof. Xue Bin Peng, and at RAI working with Dr. Farbod Farshidian, where I train policies on E-Atlas and Unitree G1.

My research lies at the intersection of computer graphics and robotics. I am interested in how humanoid agents acquire and reuse human-like motor skills to interact with the physical world and other agents. My work focuses on developing RL-based motion imitation methods to learn robust and generalizable controllers from human motion data. As a next step, I study how humanoid agents can strategically compose and coordinate behaviors over time to operate in complex, dynamic, and competitive environments.

Email / Google Scholar / Github

Students I Mentored

Kangni Liu → Fauna Robotics (acquired by Amazon)

Zihan Wang → Amazon FAR

Zehao Zhu → Ph.D. student at UT Austin

Publications

* indicates equal contributions.

Generalizing from References using a Multi-Task Reference and Goal-Driven RL Framework
Jiashun Wang, M. Eva Mungai, He Li, Jean Pierre Sleiman, Jessica Hodgins, Farbod Farshidian
RSS, 2026

paper / project page

HIL: Hybrid Imitation Learning for Dynamic Athletic Control
Jiashun Wang, Yifeng Jiang, Haotian Zhang, Chen Tessler, Davis Rempe, Jessica Hodgins, Xue Bin Peng
ACM Transactions on Graphics, 2026

paper / project page

REST3D: Reconstructing Physically Stable 3D Scenes from a Single Image
Xiaoxuan Ma, Jiashun Wang, Nicolas Ugrinovic, Yehonathan Litman, Kris Kitani
SIGGRAPH Asia, 2026

paper / code / project page

Web2Grasp: Learning Functional Grasps from Web Images of Hand-Object Interactions
Hongyi Chen, Yunchao Yao, Yufei Ye, Zhixuan Xu, Homanga Bharadhwaj, Jiashun Wang, Shubham Tulsiani, Zackory Erickson, Jeffrey Ichnowski
IROS, 2026

paper / project page

ZEST: Zero-shot Embodied Skill Transfer for Athletic Robot Control
Jean Pierre Sleiman, He Li, Alphonsus Adu-Bredu, Robin Deits, Arun Kumar, Kevin Bergamin, Mohak Bhardwaj, Scott Biddlestone, Nicola Burger, Matthew A. Estrada, Francesco Iacobelli, Twan Koolen, Alexander Lambert, Erica Lin, M. Eva Mungai, Zach Nobles, Shane Rozen-Levy, Yuyao Shi, Jiashun Wang, Jakob Welner, Fangzhou Yu, Mike Zhang, Alfred Rizzi, Jessica Hodgins, Sylvain Bertrand, Yeuhi Abe, Scott Kuindersma, Farbod Farshidian
Science Robotics, 2026

paper / video

	CRISP: Contact-guided Real2Sim from Monocular Video with Planar Scene Primitives Zihan Wang, Jiashun Wang, Jeff Tan, Yiwen Zhao, Jessica Hodgins, Shubham Tulsiani, Deva Ramanan ICLR, 2026 paper / project page
	ASAP: Aligning Simulation and Real-World Physics for Learning Agile Humanoid Whole-Body Skills Tairan He, Jiawei Gao, Wenli Xiao, Yuanhang Zhang, Zi Wang, Jiashun Wang, Zhengyi Luo, Guanqi He, Nikhil Sobanbab, Chaoyi Pan, Zeji Yi, Guannan Qu, Kris Kitani, Jessica Hodgins, Linxi "Jim" Fan, Yuke Zhu, Changliu Liu, Guanya Shi RSS, 2025 paper / project page