About

Research Interests

I am broadly interested in robotics, computer vision, and optimization and learning. In particular, my research focuses on applying rigorous (mathematical and optimization) theory to develop certifiable and provably correct perception algorithms for safety-critical applications and human-level understanding of the real world. Moving forward, I want to get more engaged in learning-based perception and apply my experience in geometric vision, robust estimation and optimization to improve the efficiency, robustness, interpretability, and theoretical guarantees of machine learning and artificial intelligence in general.

  • Robotic perception, (geometric) computer vision
  • Robust estimation
  • Semidefinite programming (SDP) and Sums-of-Squares (SOS) relaxations
  • Fast semidefinite programmig (SDP) solver
  • Computational algebraic geometry
  • Machine learing and artificial intelligence
  • Dynamical systems
  • Optimization theory

Education

  • 2018-Present, Massachusetts Institute of Technology, Cambridge, MA
    • Doctor of Philosophy (PhD), Department of Mechanical Engineering
    • Topic: robotic perception, computer vision, optimization
    • Advisor: Prof. Luca Carlone
  • 2015-2017, Massachusetts Institute of Technology, Cambridge, MA
  • 2011-2015, Tsinghua University, Beijing, China
    • Bachelor of Science (B.S.), Department of Automotive Engineering
    • Thesis: Study of Beijing Passenger Vehicle Travelling Pattern based on Big Data and Applications on Evaluation of New Energy Vehicles
    • Thesie Supervisor: Dr. Hewu Wang

Honors and Awards

  • 2021: MIT Spotlight: Making self-driving cars safer through keener robot perception (Spotlight) (Full story)
  • 2021: Best Paper Award Finalist, Robotics: Science and Systems (RSS) (Website) (Paper) (Video)
  • 2021: Robotics: Science and Systems (RSS) Pioneer (Website) (Research Statement) (Poster)
  • 2021: Best Paper Award Honorable Mention, IEEE Robotics and Automation Letters (RA-L) 
  • 2021: Finalist, MIT's first Research SLAM featuring 3MT thesis (Video)
  • 2020: Best Paper Award in Robot Vision, International Conference on Robotics and Automation (ICRA)
  • 2015: Tsinghua University Undergraduate Principal Scholarship, aka Graduation with Distinction (9/3000+)
    • 2015清华大学本科生特等奖学金​

Experience

  • Intel Intelligent Systems Lab. Research Intern, May-Sept. 2020
  • Tesla, Inc. Advanced Technology Integration Intern, June-Sept. 2017
    • New vehicle low voltage system integration.

Media Coverage

Professional Activities

Talks and Presentations

  • Certifiable Outlier-Robust Geometric Perception: Robots that See through the Clutter with Confidence

    • 2021/10: Invited Talk at the Robotics and Autonomous Systems Seminar of the Hong Kong University of Science and Technology (HKUST)

    • 2021/10: Invited Talk at the College of Computing and Informatics Research Seminar of University of North Carolina Charlotte (UNC Charlotte)

    • 2021/09: Invited Talk at Cornell Robotics Seminar

  • 2021/05: Invited Talk: Self-supervised Geometric Perception, at Imperial College London MatchLab, 2021. (Slides)

  • 2021/05: Guest Lecture: Certifiable Outlier-Robust Machine Perception, at the undergraduate course Robotics: Science and Systems (6.141J/16.405J), MIT, 2021. (Instructor: Prof. Luca Carlone) (Slides)

  • 2021/04: Guest Lecture: Certifiable Outlier-Robust Machine Perception, at the graduate course Visual Navigation for Autonomous Aerial Vehicles (VNA2V, AEROSP 740), University of Michigan, Ann Arbor, 2021. (Instructor: Prof. Vasileios Tzoumas)

  • 2021/03: A Fast Certifier for Large-Scale Degenerate SDPs in Outlier-Robust Machine Perception, at MIT LIDS & Stats Tea, 2021. 

  • 2020/09: Certifiably Robust Algorithms: From Global Optimization to Safer Perception, at MIT Driverless, 2020. (Video)

  • 2020/07: Certifiably Robust Geometric Perception with Outliers, at "Certifiable Robot Perception: from Global Optimization to Safer Robots" workshop, in conjunction with Robotics: Science and Systems (RSS), 2020. (Slides) (Video)

  • 2020/06: Invited Talk: Certifiably Robust Geometric Perception with Outliers: From TEASER to Beyond, at Marine Robotics Group at MIT. (Slides) (Video)

  • 2019/11: Hands-on Tutorial on Global Optimization in Matlab, at "Global Optimization for Geometric Understanding with Provable Guarantees", in International Conference on Computer Vision (ICCV) Tutorial, 2019. (Tutorial Download)

  • 2019/10: A Certifiably Optimal Solution for Robust Registration with Extreme Outlier Rates, at “Northeast Robotics Colloquium” (NERC), 2019. (6 selected out of 38 submissions) (Slides) (Poster)

Reviewer for International Conferences and Journals

  • Conferences: Robotics: Science and Systems (RSS); International Conference on Computer Vision (ICCV); International Conference on Robotics and Automation (ICRA); International Conference on Intelligent Robots and Systems (IROS); Conference on Computer Vision and Pattern Recognition (CVPR); Conference on Neural Information Processing Systems (NeurIPS)
  • Journals: IEEE Transactions on Robotics (T-RO), IEEE Robotics and Automation Letters (RA-L), IEEE Transactions on Pattern Analysis and Machine Intelligence (T-PAMI), Graphical Models, International Journal of Robotics Research (IJRR), Journal of Mathematical Imaging and Vision, Autonomous Robots, Computational Optimization and Applications (COAP), Journal of Field Robotics

 

[Last edit: 10/09/2021]