Peihao Zhu
San Francisco, California, United States
Summary
🤩
Rockstar🎓
Top SchoolPeihao Zhu is a Senior Research Scientist with a decade of experience at the intersection of academic research and industry productization, currently driving applied ML research at ByteDance after internships at Meta and Snap. A KAUST PhD candidate in Computer Science, he combines deep technical rigor—evidenced by publications on Google Scholar—with hands-on engineering contributions such as improving GAN-based image compositing masks for the SIGGRAPH Asia Barbershop project. He has a track record of turning research prototypes into robust pipeline components (e.g., mask inpainting, dilation/erosion enhancements) that improve real-world image processing. Based in the San Francisco Bay Area, Peihao bridges academic insight and production needs, excelling at both algorithm design and practical implementation.
10 years of coding experience
2 years of employment as a software developer
Phd Candidate, Computer Science, Phd Candidate, Computer Science at KAUST (King Abdullah University of Science and Technology)Computer Science, Computer Science at University of Chinese Academy of SciencesBachelor of Engineering in Automation, Bachelor of Engineering in Automation at Northeastern University (CN)
Chinese, English
Github Skills (12)
inpaint10
computer-vision10
pytorch10
image-segmentation10
segmentation10
python10
image-processing10
inpainting10
dcgan9
cyclegan9
machine-learning9
cgan9
Programming languages (2)
Github contributions (5)
ZPdesu/Barbershop
Jun 2021 - Apr 2022
Barbershop: GAN-based Image Compositing using Segmentation Masks (SIGGRAPH Asia 2021)
Role in this project:
ML Engineer
ML EngineerContributions:31 commits, 2 PRs, 26 pushes in 10 months
Contributions summary:Peihao primarily contributed to the project by modifying the `Alignment` module, which is a core component in the image compositing pipeline. Their changes involved implementing an auto mask inpainting process, which enhances the quality of segmentation masks. They also added dilation and erosion techniques for improved mask processing. The user also made changes to the main script, and the bash file, suggesting a contribution to the overall workflow and execution of the project.
compositingsegmentationsiggraph
ZPdesu/II2S
Jul 2021 - Apr 2022
Improved StyleGAN Embedding: Where are the Good Latents?
Contributions:22 commits, 6 pushes, 2 comments in 9 months
stylegangangenerative-modelimage-embeddingdeeo-learning
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.