About Me
Hi! I train & ship 3D visual perception models at Tesla. These models are at the core of many cool Tesla products, including Autopilot/FSD, High-Fidelity Park Assist, Vision Autopark, and Actually Smart Summon. My main engineering & research interests include computer vision foundation models, generative modeling, 3D scene understanding, and vision-language models.
Previously, I researched computer vision & deep learning at the Statistical Visual Computing Lab for 5 years. I published several papers to CVPR/CVPRW; topics include 3D reconstruction, 2D/3D detection, domain adaptation, self-supervised learning. For details, see the projects page.
I earned my Masters degree in Machine Learning and Data Science at UC San Diego. More details about my background can be found in my resume.
I also developed & maintain a set of 250+ ML Flashcards, covering computer science, classical ML, modern deep learning, 2D/3D computer vision, NLP, & reinforcement learning. This has been relatively popular & useful for others – free to check it out for reference or study!
Highlights and News
- US Patent US2025/0028326A1 Published for Tesla Vision Park Assist: “Enhanced User Interface Generation for Parking Based on Occupancy Machine Learning Models”
- Awarded Tesla Exceptional Performance Equity Grant, given to top 20% talent in Autopilot [Aug 2024]
- Promoted to Senior Machine Learning & Computer Vision Engineer at Tesla [Feb 2024]
- Machine Learning & Computer Vision Engineer at Tesla [Aug 2022]
- Masters Degree in Machine Learning and Data Science at UCSD; MS thesis studies learned visual invariances using hierarchical datasets [Apr 2022]
- Published CVPRW 2022 Paper on test-time 3D shape refinement for domain robustness [Apr 2022]
- Awarded NSF Graduate Research Fellowship (NSF GRFP) from the US National Science Foundation Agency [Mar 2020]
- Awarded Sloan Graduate Fellowship from the Sloan Foundation [Sep 2019]
- Published CVPR 2019 Paper on neural network robustness [Jun 2019]
Contact
BrandnLeung@gmail.com