Haozhi Qi
Applied Scientist at Amazon
Berkeley, California, United States
Summary
🤩
Rockstar🎓
Top SchoolHaozhi Qi is an applied scientist with 11 years of experience in computer vision and deep learning, currently working in the San Francisco Bay Area. He holds a PhD in Computer Science from UC Berkeley and has a strong research-to-production track record from internships and research roles at Microsoft and multiple visiting positions at Meta. Haozhi has hands-on expertise in instance segmentation and deformable convolutional networks, contributing to well-known open-source projects (FCIS, Deformable-ConvNets) where he fixed core bugs, added FPN support, and managed releases. His work helped win major object detection/segmentation challenges early in his career and he combines rigorous academic training with practical model deployment and maintenance skills. Colleagues would find him as comfortable debugging low-level convolutional implementations as iterating on system-level releases for production ML.
11 years of coding experience
3 years of employment as a software developer
Doctor of Philosophy - PhD Computer Science, Doctor of Philosophy - PhD Computer Science at University of California, BerkeleyHong Kong University of Science and Technology (HKUST)
Github Skills (11)
semantic-segmentation10
object-detection10
computer-vision10
mxnet10
machine-learning10
deep-learning10
instance-segmentation10
python10
tensorflow9
c-language8
cprogramming-language8
Programming languages (4)
Github contributions (5)
msracver/FCIS
May 2017 - Aug 2018
Fully Convolutional Instance-aware Semantic Segmentation
Role in this project:
ML Engineer
ML EngineerContributions:22 commits, 5 PRs, 20 pushes in 1 year 2 months
Contributions summary:Haozhi made several releases, indicating a focus on preparing and deploying the project. They addressed mask merging issues, a key component of instance segmentation. Additionally, the user updated the license information, suggesting an active role in managing and maintaining the project's core functionalities and distribution.
semantic-segmentationbackbonedeep-learningcomputer-visionconvolutional
msracver/Deformable-ConvNets
May 2017 - Jun 2018
Deformable Convolutional Networks
Role in this project:ML Engineer
ML EngineerContributions:20 commits, 6 PRs, 17 pushes in 1 year 1 month
Contributions summary:Haozhi contributed to the development and maintenance of a deep learning project focused on deformable convolutional networks. Their work included cleaning up helper functions, fixing bugs within the core deformable convolution implementation, and adding support for feature pyramid networks (FPN), a common architectural component in object detection models. Further contributions involved updating the deformable convolution implementation.
efficientnetdeep-learningconvolutionalconvolutional-networksresnet
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.