Wenxuan Zhou

I am a Ph.D. student at the Robotics Institute (RI) at Carnegie Mellon University, advised by Prof. David Held. My research goal is to enable task-level autonomy for robots through data-driven approaches such as reinforcement learning. I will be joining DeepMind as an intern in Summer 2021.

I received my master's degree at RI mentored by Lerrel Pinto and advised by Prof. Abhinav Gupta. During my undergraduate study, I worked with Prof. Gabor Orosz on ground robot experiments with connected cruise control. I've also interned at ZF TRW at the brake control systems group.

Contact  /  GitHub  /  Google Scholar

profile photo
Research
clean-usnob PLAS: Latent Action Space for Offline Reinforcement Learning
Wenxuan Zhou, Sujay Bajracharya, David Held
CoRL 2020 (Plenary Talk)

Learning policy in the latent action space to naturally avoid out-of-distribution actions.

#Offline_RL #Off_Policy_RL #Cloth_Manipulation

[Paper] [Code] [Website]
clean-usnob Lyapunov Barrier Policy Optimization
Harshit Sikchi, Wenxuan Zhou, David Held
In submission

#Safe_RL

clean-usnob Learning Off-Policy with Online Planning
Harshit Sikchi, Wenxuan Zhou, David Held
ICML 2020 Workshop on Inductive Biases, Invariances and Generalization in RL

Enhancing model-based CEM with a terminal Q-function learned by an off-policy algorithm.

#Model_Based_RL #Model_Free_RL

[Paper]
clean-usnob EPI: Environment Probing Interaction Policies
Wenxuan Zhou, Lerrel Pinto, Abhinav Gupta
ICLR 2019

Learning to "probe" the environment before task execution.

#System_Identification #Environment_Generalization #Multi_Task_Learning

[Paper][Code]

Source