Shuyang Sun
Senior Research Engineer at Google DeepMind
United States
Summary
👤
Senior🎓
Top SchoolShuyang Sun is a Senior Research Engineer with a decade of experience specializing in computer vision and segmentation research, currently at Google DeepMind. He earned advanced degrees from Oxford and the University of Sydney and has a track record of leading projects that set state-of-the-art results (NeurIPS 2023, CVPR 2024) on efficient panoptic and open-vocabulary segmentation. His contributions span both research and engineering: from designing core ideas and end-to-end pipelines to implementing production-ready model refinements, including notable work integrating SAM and improving CamCut within the prominent google-research repository. Known for translating novel ideas into reproducible code and papers, he has moved rapidly through research roles at Google, ByteDance, Intel, and academia, combining deep theoretical insight with practical system-building. An oft-overlooked strength is his ability to boost baseline performance substantially without heavy retraining, demonstrating a knack for training-free and efficiency-focused solutions.
10 years of coding experience
3 years of employment as a software developer
BEng Software Engineering, BEng Software Engineering at Wuhan UniversityDoctor of Philosophy in Engineering Science Computer Vision, Doctor of Philosophy in Engineering Science Computer Vision at University of OxfordThe University of Sydney
Github Skills (8)
computer-vision10
pytorch10
machine-learning10
image-segmentation10
segmentation10
clip10
python10
ai10
Programming languages (4)
Github contributions (5)
google-research/google-research
Feb 2024 - Jul 2024
Google Research
Role in this project:
ML Engineer
ML EngineerContributions:11 PRs in 4 months
Contributions summary:Shuyang contributed to the implementation and refinement of the CamCut model, specifically focusing on image segmentation tasks within the Google Research repository. Their work involved modifying the CamCut model for semantic segmentation, developing functions for visual prompting and confidence scoring, and integrating SAM for enhanced mask generation. The commits show a strong focus on adapting and improving the model's performance, particularly in refining mask filtering and integration with other segmentation techniques.
googlemachine-learningai
OOP-JAVA-WHUISS/Assignment2
Oct 2015 - Jan 2016
Contributions:134 PRs, 133 pushes, 1 branch in 2 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.