Zhiwen Chen
Ph.D. Student, HKUST Research Assistant, SJTU UniAD Algo Intern, SenseTime B.Eng. Vehicle Engineering, Tongji
Hi! 👋 I'm Zhiwen Chen, currently a Ph.D. student at the Hong Kong University of Science and Technology, advised by Prof. Jun Ma, and a Research Assistant at Shanghai Jiao Tong University, advised by Prof. Zhipeng Zhang. I received my B.Eng. 🎓 in Vehicle Engineering from Tongji University. I also serve as a reviewer for NeurIPS, CVPR, ECCV, among others.
My research sits at the intersection of 🚗 End-to-End Autonomous Driving, 🤖 Reinforcement Learning, and 🔮 Large Vision-Language Models. I focus on RL- and LLM-driven end-to-end driving and trajectory planning, aiming to build agentic driving systems that coordinate multiple planners and make human-like decisions in complex urban environments. 🏙️
Timeline
Ph.D. Student
Hong Kong University of Science and Technology
Research Assistant
Shanghai Jiao Tong University
Graduate Researcher, Vehicle Engineering
Tongji University
Algorithm Research Intern
SenseTime Research
Short-term Intern
SAIC Volkswagen
B.Eng. in Vehicle Engineering
Tongji University
Publications
KDD 2026
CCF A
Oral
Tongji University
The Hong Kong University of Science and Technology (Guangzhou)
The Hong Kong Polytechnic UniversityA. Wang, Zhiwen Chen, S. Wang, Q. Xia, J. Li
NeurIPS 2025
CCF A
Accepted
Tongji University
SenseTime ResearchZhiwen Chen, H. Deng, Z. Li, H. Wen, G. Jin, R. Yu, B. Leng
Under Review
FlexPlanner: Learning Adaptive Driving Planner with Temporal Reward Decomposition
The Hong Kong University of Science and Technology (Guangzhou)
Shanghai Jiao Tong University
Hello Inc.Under Review
RegAD: Rethinking Regressive versus Generative Planning in End-to-End Autonomous Driving
IEEE TPAMI 2026
Under Review
Typed Budgets for Object-Relational Driving World Models
Under Review
Survey of General End-to-End Autonomous Driving: A Unified Perspective
SJTU
DiDi ChuxingIEEE Transactions on Neural Networks and Learning Systems
Under Review
A Survey of Reinforcement Learning-Based Motion Planning for Autonomous Driving: Lessons Learned from a Driving Task Perspective
Tongji University
University of Michigan
Texas A&M UniversityZ. Li, G. Jin, R. Yu, Zhiwen Chen, N. Li, W. Han, L. Xiong, B. Leng, J. Hu, I. Kolmanovsky, D. Filev
Journal of Tongji University (Natural Science Edition) 2022
Accepted
A Global Path Planning Method for Multi-Vehicle Autonomous Parking Based on Multi-Attribute Decision Making
Tongji UniversityZhiwen Chen, Z. Li, J. Wu, et al.
Experience
Research Assistant — Shanghai Jiao Tong University
Conducting research on automated driving motion planning methods at the Artificial Intelligence Institute. Focusing on utilizing reinforcement learning (RL) and world models instead of Mixture-of-Experts (MoE) techniques in LLMs.
Algorithm Research Intern — SenseTime Research
Built scenarios in the CARLA simulator to generate simulation data for the UniAD framework. Performed subsequent data processing and model training based on the generated simulation data.
Short-term Intern — SAIC Volkswagen
Gained an understanding of the EHH department's structure. Explored the applications and technical prospects of AI in intelligent driving.
Portfolio
Roadside-Information-Enhanced Interactive Learning for Decision-Making and Control
Joint project with Changan Automobile. Constructed risk constraints against unreasonable driving actions. Established a self-learning decision and control method guided by rule-based mechanisms and game theory.
Research on Planning and Control Algorithms for Collaborative Autonomous Parking
Optimized local parking paths and completed validation through real-world vehicle testing.
A Global Path Planning Platform for Multi-Vehicle Parking
Developed a global path planning platform for multi-vehicle parking. Implemented a complete end-to-end automatic parking system covering multi-vehicle space allocation, global guidance, local path planning, and demonstration on a visualization platform.
An Intelligent Sign Language Translation and Interaction System for the Hearing-Impaired
Joint project with Huawei. Achieved real-time recognition of sign language and conversion into text.