Tianmin Shu

I am Tianmin Shu, a fouth year Ph.D. student in the Department of Statistics at University of California, Los Angeles. I currently work at the Center for Vision, Cognition, Learning, and Autonomy (VCLA) under the supervision of Prof. Song-Chun Zhu. I also closely work with Prof. Sinisa Todorovic, Prof. Hongjing Lu and Prof. Michael S. Ryoo.

Before coming to UCLA, I spent four years at Fudan University, where I received my bachelor's degree in Electronic Engineering. I was an exchange student at UCLA in the summer of 2013.

Google Scholar Profile


08/2017: Invitated talk at ONR MURI meeting, UCLA.

07/2017: Invitated talk at 3rd Vision Meets Cognition Workshop in conjunction with CVPR 2017, Honolulu, HI.

06/2017: I started a deep learning research intenrship at Salesforce Research (MetaMind)

06/2017: Our CogSci 2017 paper received Computational Modeling Prize in Perception and Action.

04/2017: Our ICRA 2017 work was covered by New Scientist. [Link]   

Reseach Highlights

Human-Robot Interactions

Transferring human interactions to human-robot interactions (ICRA 2017) Learning social affordances and synthesizing human skeleton sequences of social interactions (IJCAI 2016)

Human Activity Recognition

Energy based RNNs regularized by confidence for group activity recogntion (CVPR 2017) Aerial video analysis (CVPR 2015) Human trajectory prediction (PAMI 2017)

Computational Cognitive Science

Inferring human interaction from motion trajectories (CogSci 2017) Critical features of human interaction (CogSci 2016)