Feiyang Ma
Machine Learning Researcher, Efficient Accelerationist
I am a Master student in the Department of Computer Science and Engineering at the University of California, San Diego. I finished my undergraduate study at University of California, Irvine majoring in Computer Science with a specialization in intelligence systems. My research interests are in the field of computer vision and machine learning. My goal is to build spatial intelligence systems that can operate in the real world and generalize to unseen scenarios. My research interests include:
- Self-supervised learning in 3D vision
- Geometric deep learning for building strong priors
I'm currently learning about group representation theory including SE(3) equivalence and how it could benefit learning. In my spare time, I enjoy reading books, going to gym, playing video games and soccer. In fact, I might join the volunteering group of FIFA World Cup in Los Angeles in June 2026! If you are part of the soccer world you should try to sign up for the tryout! This is the La Copa de la Vida! Let's see if Argentina wins the World Cup again!
Selected Publications (full list)
Self-Supervised Learning with 2D-3D Correspondences in Autonomous Driving
Authors: Feiyang Ma, Co-author One, Co-author Two
Conference on Computer Vision and Pattern Recognition (CVPR), 2026
Geometric Priors in Action Learning for Robot Manipulation
Authors: Feiyang Ma, Another Author
International Conference on Machine Learning (ICML), 2026
Experience
Working on self-supervised learning for 3D vision tasks.
- Developed novel architectures for processing point cloud data.
- Researched applications of geometric deep learning and Lie groups.
- Prepared manuscripts for publication in top-tier conferences.
Implemented and tested production models for 3D reconstruction.
If you have any questions, please feel free to contact me via email at feiyangma@ucsd.edu.