Yunhao Fang
Research Scientist at Physical Intelligence
California, United States
Summary
🤩
Rockstar🎓
Top SchoolYunhao Fang is a research scientist based in California with six years of experience building and improving multimodal and video perception systems. He has driven core development on open-source video tracking (mmtracking) at Shanghai AI Lab and improved real-world evaluation by adding HOTA metrics and TAO dataset support, demonstrating attention to robust benchmarking. His industry work includes research roles at NVIDIA and ByteDance focused on multimodal self-evolving systems and SEED multimodal research, and he recently joined Physical Intelligence to apply those skills. Trained at Zhejiang University and UC San Diego (MS CS), he blends strong academic foundations with hands-on engineering that fixes platform bugs, cross-platform installs, and subtle tracker data-type issues. Known on GitHub as a "free and curious mind," he favors practical reliability improvements that make research code production-ready.
5 years of coding experience
1 year of employment as a software developer
University of California, San DiegoBachelor of Engineering - BE Information Science and Electronic Engineering, Bachelor of Engineering - BE Information Science and Electronic Engineering at Zhejiang University
Github Skills (5)
multi-object-tracking10
computer-vision10
python10
pytorch9
git8
Programming languages (1)
Github contributions (5)
open-mmlab/mmtracking
Dec 2021 - Mar 2022
OpenMMLab Video Perception Toolbox. It supports Video Object Detection (VID), Multiple Object Tracking (MOT), Single Object Tracking (SOT), Video Instance Segmentation (VIS) with a unified framework.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:7 reviews, 14 commits, 42 PRs in 3 months
Contributions summary:Yunhao contributed to improving the project by addressing installation issues, fixing bugs, and enhancing the codebase. They implemented symbolic link creation on Windows, integrated the MMHuman3D link, and addressed a data type bug in ByteTracker, improving the reliability of the tracking algorithms. The user also added support for HOTA evaluation metrics and integrated TAO dataset, enhancing evaluation capabilities and broader dataset support.
sotvideovisopenmmlabmultiple-object-tracking
Seerkfang/mmtracking
Dec 2021 - May 2022
OpenMMLab Video Perception Toolbox. It supports Video Object Detection (VID), Multiple Object Tracking (MOT), Single Object Tracking (SOT), Video Instance Segmentation (VIS) with a unified framework.
Contributions:99 pushes, 33 branches in 5 months
sotvideovisopenmmlabmultiple-object-tracking
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.