Lei TAI 邰磊

Ph.D. candidate

Robotics Institute, Department of Electronic and Computer Engineering

The Hong Kong University of Science and Technology

Email: ltai AT ust DOT hk


About Me

I am a fifth-year Ph.D. candidate in RAM Lab, The Hong Kong University of Science and Technology (HKUST), supervised by Prof. Ming Liu. I received the B.S.(2012) and the M.S.(2014) in Engineering from Harbin Institute of Technology (HIT). In the last year of the undergraduate, I joined the ABU Robocon as a team member of HIT Competitive Robot Team. Since then, I started my robotics work. From March 2017 to Jan 2018, I was a visiting researcher in AIS lab, University of Freiburg, working with Prof. Joschka Boedecker and Prof. Wolfram Burgard. My research interests include deep reinforcement learning and learning from demonstrations on mobile robots.

[google scholar] [github] [CV]


Research Publications

Sensorimotor learning

  • VR-Goggles for Robots: Real-to-sim Domain Adaptation for Visual Control.
    Jingwei Zhang*, Lei Tai*, Yufeng Xiong, Peng Yun, Ming Liu, Joschka Boedecker, Wolfram Burgard
    (* indicates equal contribution)
    arxiv 1802.00265
    arXiv video

  • Curiosity-driven Exploration for Mapless Navigation with Deep Reinforcement Learning.
    Oleksii Zhelo, Jingwei Zhang, Lei Tai, Ming Liu, Wolfram Burgard
    ICRA 2018 Workshop on Machine Learning in Planning and Control of Robot Motion.
    page pdf bibtex video

  • Socially Compliant Navigation through Raw Depth Inputs with Generative Adversarial Imitation Learning.
    Lei Tai, Jingwei Zhang, Ming Liu, Wolfram Burgard
    International Conference on Robotics and Automation (ICRA), 2018.
    link pdf bibtex dataset video

  • Neural SLAM: Learning to Explore with External Memory.
    Jingwei Zhang, Lei Tai, Joschka Boedecker, Wolfram Burgard, Ming Liu
    arxiv 1706.09520
    arXiv video

  • Virtual-to-real Deep Reinforcement Learning: Continuous Control of Mobile Robots for Mapless Navigation.
    Lei Tai, Giuseppe Paolo, Ming Liu
    IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2017.
    link pdf bibtex video
  • A Survey of Deep Network Solutions for Learning Control in Robotics: From Reinforcement to Imitation.
    Lei Tai*, Jingwei Zhang*, Ming Liu, Joschka Boedecker, Wolfram Burgard
    (* indicates equal contribution)
    arxiv 1612.07139
    arXiv

  • Towards Cognitive Exploration through Deep Reinforcement Learning for Mobile Robots.
    Lei Tai, Ming Liu
    arxiv 1610.01733
    arXiv

  • A Robot Exploration Strategy Based on Q-learning Network.
    Lei Tai, Ming Liu
    IEEE International Conference on Real-time Computing and Robotics (RCAR) 2016.
    link pdf bibtex

  • A Deep-network Solution Towards Model-less Obstacle Avoidance.
    Lei Tai, Shaohua Li, Ming Liu
    IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), 2016.
    link pdf bibtex dataset

3D Perceptions

  • Focal Loss in 3D Object Detection.
    Peng Yun, Lei Tai, Yuan Wang, Ming Liu
    arxiv 1809.06065
    arXiv code

  • PointSeg: Real-Time Semantic Segmentation Based on 3D LiDAR Point Cloud.
    Yuan Wang, Tianyue Shi, Peng Yun, Lei Tai, Ming Liu
    arxiv 1807.06288
    arXiv video code

All the research videos are also available at [bilibili].


Academic Service

  • Program Committee Member: ICVS 2017, RCAR 2016
  • Conference Reviewer: ICRA 2017-19, IROS 2016-18
  • Journal Reviewer: Trans-NNLS, IJARS, Robotics and Automation Letters

Misc