Zhejun Zhang

prof_pic.JPG

I am a research internship at the NVIDIA autonomous vehicle research group led by Prof. Marco Pavone, and a PhD student advised by Prof. Luc Van Gool at the Computer Vision Lab (CVL) of ETH Zurich. My PhD project is funded by Toyota Research on Automated Cars in Europe (TRACE, TRACE Zurich). Before joining CVL I received my BSc in ITET from TU Munich and my MSc in ITET from ETH Zurich. As master thesis topic I worked on vision-based control and considered that as my research interest since then.

My current research interests include imitation learning, reinforcement learning, world models, Transformers and their applications to autonomous driving and robotics. In particular, I tackle the problem of planing in highly interactive driving scenarios from a data-driven perspective by combining end-to-end policy learning and neural simulation.

selected publications

  1. Real-Time Motion Prediction via Heterogeneous Polyline Transformer with Relative Pose Encoding
    Zhejun ZhangAlexander LinigerChristos Sakaridis, Fisher Yu, and Luc Van Gool
    In Advances in Neural Information Processing Systems (NeurIPS) 2023
  2. TrafficBots: Towards World Models for Autonomous Driving Simulation and Motion Prediction
    Zhejun ZhangAlexander LinigerDengxin Dai, Fisher Yu, and Luc Van Gool
    In International Conference on Robotics and Automation (ICRA) 2023
  3. A Multiplicative Value Function for Safe and Efficient Reinforcement Learning
    Nick BührerZhejun ZhangAlexander Liniger, Fisher Yu, and Luc Van Gool
    In International Conference on Intelligent Robots and Systems (IROS) 2023
  4. End-to-End Urban Driving by Imitating a Reinforcement Learning Coach
    Zhejun ZhangAlexander LinigerDengxin Dai, Fisher Yu, and Luc Van Gool
    In Proceedings of the IEEE/CVF International Conference on Computer Vision (ICCV) 2021