Yu Sun is a research engineer with seven years of experience specializing in monocular 3D human mesh recovery and pose estimation. He has moved from industry research internships into applied ML roles, most recently at Meshcapade and now at Epic Games, translating state-of-the-art academic models into production-ready systems. His GitHub contributions to ROMP show practical impact on multi-person 3D regression, including model redesign, removing heavyweight dependencies, and improving real-time webcam and visualization demos. Holding a PhD from Harbin Institute of Technology, he blends deep academic training with hands-on engineering, optimizing models for deployment. Based in China, Yu is skilled at bridging research advances and product constraints to deliver robust, efficient 3D perception pipelines.
Monocular, One-stage, Regression of Multiple 3D People and their 3D positions & trajectories in camera & global coordinates. ROMP[ICCV21], BEV[CVPR22], TRACE[CVPR2023]
Role in this project:
ML Engineer
Contributions:11 releases, 285 commits, 16 PRs in 2 years 3 months
Contributions summary:Yu primarily contributed to the development and maintenance of a monocular 3D human pose and mesh recovery system. Their work involved uploading and integrating a new model file (`modelv5.py`), likely improving the system's core functionality. They also made changes related to saving mesh results and removed dependencies on a related library (`smplx`), indicating a focus on optimizing the system's performance or architecture. Finally, the commits demonstrate that they are working on webcam demo and multi-person visualization.
Contributions:69 commits, 45 pushes, 1 comment in 1 year 5 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.