Shiyu Duan
Machine Learning Engineer at Apple
Cupertino, California, United States
Summary
👤
Senior🎓
Top SchoolShiyu Duan is a machine learning engineer at Apple with nine years of experience applying deep learning to image and video problems, grounded by a Ph.D. in Electrical and Computer Engineering from the University of Florida. His doctoral work focused on foundational deep learning theory under Dr. Jose Principe, giving him a strong theoretical lens for practical system design. At Apple he builds ML solutions for visual media, and his open-source contributions include implementing a PyTorch Projected Gradient Descent attack in the well-known CleverHans adversarial example library. He has industry research experience from internships at SenseTime and Bosch, including GAN-based image/video synthesis and analytics for healthcare recovery, reflecting a knack for translating research into applied systems. Based in Cupertino, he combines academic rigor with production-oriented engineering, often bridging adversarial robustness and video/image processing in real-world products.
9 years of coding experience
3 years of employment as a software developer
Bachelor of Science - BS, Electrical Engineering, Bachelor of Science - BS, Electrical Engineering at Fudan UniversityDoctor of Philosophy - PhD, Electrical and Computer Engineering, Doctor of Philosophy - PhD, Electrical and Computer Engineering at University of Florida
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (7)
pytorch10
adversarial-machine-learning10
machine-learning10
python10
security9
benchmark9
benchmarking9
Programming languages (2)
Github contributions (5)
cleverhans-lab/cleverhans
Apr 2019 - May 2020
An adversarial example library for constructing attacks, building defenses, and benchmarking both
Role in this project:
ML Engineer
ML EngineerContributions:68 commits, 9 PRs, 70 comments in 1 year 1 month
Contributions summary:Shiyu implemented a PyTorch-based implementation of the Projected Gradient Descent (PGD) attack, a crucial component for adversarial example generation. The commits demonstrate an understanding of the original PGD method, its parameters, and its application within the context of adversarial machine learning. Further, the user refactored the code by moving functionality to utility modules and added examples to demonstrate the attack and perform adversarial training.
benchmarkingrobustnessadversarial-machine-learningsecurityadversarial
michaelshiyu/cleverhans
Apr 2019 - Jan 2021
An adversarial example library for constructing attacks, building defenses, and benchmarking both
Contributions:87 pushes, 16 branches in 1 year 10 months
benchmarkingadversarialadversarial-exampleexample-libraryattacks
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.