Zhicheng Wu
Research Scientist at Meta
San Jose, California, United States
Summary
👤
Senior🎓
Top SchoolZhicheng Wu is a research scientist in the San Francisco Bay Area with eight years of experience applying ML to real-world problems across industry and academia. He has worked on large-scale recommendation systems and ads integrity at Meta and TikTok, and previously built novel optical simulation and end-to-end co-optimized sensing systems during a PhD at University of Wisconsin–Madison. His background spans computer vision for medical diagnostics, parallelized ML-enabled physics simulations, and practical production pipelines, bridging deep research with deployable engineering. Notably, he has translated wave and diffraction physics into ML features for tasks like sound localization and developed platforms that focus light for neural inference—demonstrating a rare mix of optics, physics, and ML systems expertise.
8 years of coding experience
6 years of employment as a software developer
high school, high school at 包头市第一中学Doctor of Philosophy - PhD Electrical and Electronics Engineering, Doctor of Philosophy - PhD Electrical and Electronics Engineering at University of Wisconsin-MadisonBachelor of Science - BS Physics, Bachelor of Science - BS Physics at Peking University
Github Skills (22)
inverse10
periodic10
meta9
factor9
approximation9
mnist-dataset8
gradient-descent7
mnist6
classify6
classifier6
optimization5
solver5
simulation5
python4
deep-learning3
Programming languages (2)
Github contributions (5)
airPeter/Meta_SCMT
Feb 2022 - Dec 2022
Meta_SCMT is a software to inverse design large-area dielectric metasurface. The key factor that differentiates the Meta_SCMT from current Local Periodic Approximation(LPA) based methods is that Meta_SCMT can achieve fullwave-level accuracy with much less computational resource.
Contributions:82 commits, 2 PRs, 13 pushes in 9 months
accuracymetasurfacemetafactorcoupled-mode-theory
airPeter/SmartGlass
Apr 2022 - May 2022
SmartGlass (SG) is a python implementation of a diffractive optical neural network. Currently, it supports training an all-optical classifier (e.g. classify hand-written digits MNIST dataset). Besides, the framework can also be used to design optics based on a task like focusing and beam steering. However, custom object functions should be defined. Besides training the optics, the SG also supports training the detectors using gradient-free optimization.
Contributions:12 commits, 2 PRs, 10 pushes in 24 days
pythonclassifierbeammetasurfacegradient-free-optimization
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.