Ke Guo

Ke Guo

PhD of Artificial Intelligence

The University of Hong Kong (HKU)

Biography

I received the PhD degree in computer science from the University of Hong Kong, supervised by Prof. Jia Pan and Prof. Wenping Wang. I draw ideas in optimization, control and artificial intelligence to develop algorithms for autonomous driving and intelligent transportation. I currently focus on developing a learning-based traffic model.

Interests
  • Artificial Intelligence
  • Intelligent Transportation
  • Autonomous Driving
Education
  • PhD in Computer Science, 2019-2023

    The University of Hong Kong (HKU)

  • BSc in Automation, 2015-2019

    Zhejiang University (ZJU)

Recent Publications

(2024). LASIL: Learner-Aware Supervised Imitation Learning For Long-term Microscopic Traffic Simulation. In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

PDF Cite Code Video

(2023). TraCo: Learning Virtual Traffic Coordinator for Cooperation with Multi-Agent Reinforcement Learning. In Conference on Robot Learning (coRL).

PDF Cite Video DOI

(2023). CCIL: Context-conditioned imitation learning for urban driving. In {Robotics:} Science and Systems (RSS).

PDF Cite Code Video DOI

(2022). End-to-End Trajectory Distribution Prediction Based on Occupancy Grid Maps. In IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

PDF Cite Code Video DOI

(2021). VR-ORCA: Variable Responsibility Optimal Reciprocal Collision Avoidance. In IEEE Robotics and Automation Letters (RAL).

PDF Cite Code Video DOI

Research

*
Imitation Learning for Traffic simulation
Learning a driving model for traffic simulation by imitating from real-world traffic data.
Imitation Learning for Urban Driving
Learning a context-conditioned autonomous driving policy by imitating from human driver’s demonstration.
Trajectory Distribution Prediction
Predicting an multi-modal trajectory distribution for pedestrians and vehicles based on occupancy grid maps.
Decentralized Collision Avoidance
Navigating multiple agents to their goals without mutual communication and collisions.