Biography
I am now a third-year PhD student at University of Hong Kong and HKU-MMLab, fortunately supervised by Prof. Ping Luo and co-advised by Prof. Wenping Wang and Xiaoou Tang. I work closely with Dr. Wenqi Shao and Dr. Bo Zhang. My current research focus is on unsupervised 3D scene understanding and its application in autonomous driving and embodied AI.
Previously, I received my bachelor’s degree in Automation at Zhejiang Univerisity with an honor degree for the Mixed Class at Chu Kochen Honor College in July, 2020. At ZJU, I worked under the supervision of Prof. Rong Xiong and Yue Wang at ZJU.
Interests
- Unsupervised Representation Learning
- Autonomous Driving
- Long-horizon Manipulation Learning
Education
BE in Automation, 2016--2020
Zhejiang University
PhD in Computer Science, 2021--now
University of Hong Kong
Visiting Student, 2024--now
Yale University
News
- [Top] We are looking for self-motivated students to collaborate on unsupervised 3D scene understanding for autonomous driving and embodied AI (potentially with financial support). Drop me an email if you are interested.
- 2024-03-18: Our work CurriculumLoc on place recognition for remote sensing is accepted in TGRS. Congrats on Boni! [paper] [code]
- 2024-01-07: Join Yale Vision Laboratory as a visiting student.
- 2023-07-29: An extended version of CycleMLP is accepted in T-PAMI. Congrats on Shoufa! [paper] [code]
- 2023-02-28: Our work MV-JAR on MAE-based unsupervised 3D representation learning for LiDAR point clouds is accepted in CVPR 2023. Congrats on Runsen! [paper] [code]
- 2023-01-21: Our work CO^3 on Cooperative-Contrastive-based unsupervised 3D representation learning for LiDAR point clouds is accepted in ICLR 2023. Paper is available here and code is also released.
- 2023-01-17: Our work “Failure-Aware Policy Learning for Self-Assessable Robotics Tasks” is accepted in ICRA 2023. Congrats on Kechun! [paper]
- 2022-05-15: Our work CtrlFormer is accepted in ICML 2022. Congrats on Yao! [code]
- 2022-03-02: Our work RestoreFormer is accepted in CVPR 2022. Congrats on Zhouxia! [paper] [code]
- 2022-01-29: Our work CycleMLP is accepted in ICLR 2022 as oral presentation (1.6% acceptance rate). Congrats on Shoufa! [paper] [code]
- 2021-07-30: Y S and Christabel Lung Postgraduate Scholarship offer.
- 2021-04-12: Hong Kong PhD Fellowship offer.
- 2021-02-18: A paper is accepted in IEEE Transactions on Intelligent Transportation Systems. (IF=6.3) [paper]
- 2021-02-09: HKU Presidential Scholarship offer.
- 2021-02-04: An RA-L paper is accepted and will be presented in ICRA 2021. See you in Xi’an, China! [paper] [code]
Selected Publications
Quickly discover relevant content by filtering publications.
(2022).
CtrlFormer: Learning Transferable State Representation for Visual Control via Transformer.
Proceedings of the 39 th International Conference on Machine Learning.
(2022).
CycleMLP: A MLP-like Architecture for Dense Prediction.
The 10th International Conference on Learning Representations (Oral Presentation, 1.6% acceptance rate).
(2022).
Deep Samplable Observation Model for Global Localization and Kidnapping.
IEEE ROBOTICS AND AUTOMATION LETTERS (RA-L).
Awards
- Y S and Christabel Lung Postgraduate Scholarship, 2021
- Hong Kong PhD Fellowship offer, 2021
- HKU Presidential PhD Scholarship, 2021
- SUPCON Scholarship, 2019
- Second-class Scholarship for Academic Excellence, 2019
- Hailiang Scholarship, 2018
- First-class Scholarship for Academic Excellence, 2017 & 2018