Tianyu Xiang

Postdoctoral Researcher, Penn State Neuromechanics Laboratory, The Pennsylvania State University

I received my Ph.D. in Pattern Recognition and Intelligent Systems from the Institute of Automation, Chinese Academy of Sciences in June 2026. I am now conducting postdoctoral research at the Penn State Neuromechanics Laboratory at University Park. My research asks two connected questions: How do humans perform manipulation? and How can robots learn those skills efficiently? I combine video, EEG, and EMG to model human manipulation from coarse behavior to fine neural control, then translate these representations and human learning mechanisms into sample-efficient robot-learning algorithms.

Research overview: from multimodal human manipulation representation and online neural decoding to brain-inspired human–robot skill transfer.

Research highlights

Robot skill learning remains limited by low data efficiency and weak generalization. Inspired by human learning, I investigate how the brain controls muscles during skill execution and acquisition, build brain–muscle modulation models, and transfer these mechanisms to robots to enable brain-inspired skill learning.

• Human skill modeling: Modeling human manipulation and learning across video, EEG, and EMG, from observable behavior to brain–muscle control.
• Brain-inspired robot learning: Transferring human skill representations and learning mechanisms to robots for more efficient and generalizable skill acquisition.
• Research outputs: 19 first/co-first-author papers published or accepted—eight journal papers and eleven international conference papers, including six IEEE Transactions papers—and five invention-patent applications as lead drafter, with three granted.
• Recognition and support: Recipient of the 2026 Chinese Academy of Sciences President’s Special Award and the IEEE RCAR Best Paper Award; supported by the Direct Ph.D. Pilot Program, an IEEE CIS Graduate Student Research Grant, and an IEEE CIS Travel Grant.

Explore my research → · Read my biography →