詹锟 | Kun Zhan

🔭 About Me

I am Kun Zhan, Cognitive Intelligence Lead at Li Auto and Site Manager of the company’s new U.S. R&D Center.

At Li Auto, I own the Vision-Language-Action (VLA) algorithm roadmap—directing perception, planning, large-scale model training, and vehicle-grade deployment. My work has advanced from end-to-end driving (E2E) and vision-language models (VLM) to the full VLA stack, and I lead world-model research that powers end-to-end simulation, synthetic-data generation, and reinforcement learning.

My mission is to close the gap between breakthrough AI and everyday mobility. By translating cutting-edge computer vision, 3D perception, multimodal foundation models, and reinforcement learning into production vehicles and robots, I strive to deliver safer, smarter, truly autonomous transportation.

🌟 Research Interests

• Autonomous Driving: Vision-Language-Action(VLA) Model, End-to-end autonomous driving systems, decision-making and planning
• Computer Vision: Object detection and tracking, scene understanding
• 3D Vision: 3D perception, reconstruction, and modeling
• Large Language Models: Applications of multimodal large models in autonomous driving
• World Models: Environment modeling and prediction, reinforcement learning

💼 Work Experience

Li Auto | Apr 2021 – Present

Cognitive Intelligence Lead, VLA Algorithm Owner

• Direct a more than 100-person org covering perception, planning, foundation-model training, and on-vehicle inference.
• Delivered three generations of AI stacks (E2E → VLM → VLA) into mass-production pipelines.
• Established and lead a dedicated world-model group for reinforcement learning and closed-loop simulation.
• Previously headed the Prediction, Static BEV, Driving Perception, and Large-Model teams.

Site Manager, U.S. R&D Center (San Jose, CA)

• Built Li Auto’s overseas research hub.
• Own local strategy, budgeting, and cross-border collaboration with Beijing HQ, accelerating global talent acquisition and technology transfer.

Baidu | Apr 2016 – Mar 2021

Autonomous-Driving Engineering, Apollo L4 Team

• Served as Algorithm Lead for Prediction; designed L4 prediction pre-decision algorithms for robo-taxi operations.
• Developed planning-and-control (PnC) algorithms and deep-learning onboard modules for L4 pilots.

📚 Academic Achievements

Citation Statistics

• Top Papers: 32
• Total Citations: 720
• h-index: 12
• i10-index: 12

Selected Publications

1. Drivevlm: The convergence of autonomous driving and large vision-language models (2024) X Tian, J Gu, B Li, Y Liu, Y Wang, Z Zhao, K Zhan, P Jia, X Lang, H Zhao arXiv preprint arXiv:2402.12289 | Citations: 255

2. Street gaussians: Modeling dynamic urban scenes with gaussian splatting (2024) Y Yan, H Lin, C Zhou, W Wang, H Sun, K Zhan, X Lang, X Zhou, S Peng European Conference on Computer Vision, 156-173 | Citations: 211

3. Planagent: A multi-modal large language agent for closed-loop vehicle motion planning (2024) Y Zheng, Z Xing, Q Zhang, B Jin, P Li, Y Zheng, Z Xia, K Zhan, X Lang, D Zhao arXiv preprint arXiv:2406.01587 | Citations: 31

4. Unleashing generalization of end-to-end autonomous driving with controllable long video generation (2024) E Ma, L Zhou, T Tang, Z Zhang, D Han, J Jiang, K Zhan, P Jia, X Lang, K Yu arXiv preprint arXiv:2406.01349 | Citations: 27

5. Tod3cap: Towards 3d dense captioning in outdoor scenes (2024) B Jin, Y Zheng, P Li, W Li, Y Zheng, S Hu, X Liu, J Zhu, Z Yan, H Sun, K Zhan, X Lang, P Jia European Conference on Computer Vision, 367-384 | Citations: 22

Patents

• 16 Chinese Patents
• 2 US Patents

Academic Service

• Program Committee/Reviewer: CVPR, ICCV, ECCV, NeurIPS, AAAI
• Journal Reviewer: IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI), IEEE Transactions on Intelligent Transportation Systems (T-ITS), IEEE Transactions on Intelligent Vehicles (T-IV)
• Workshop Organizer: Autonomous Driving Workshop at CVPR 2023