Wuxun Zhang
Software Engineer at intel
Shanghai, China
Summary
🤩
RockstarWuxun Zhang is a software engineer based in Shanghai with 8 years of experience specializing in deep learning infrastructure and model optimization. He is an active open-source contributor to prominent projects like Apache MXNet and GluonCV, where he improved C++ core operators, integrated MKL-DNN for 3D convolutions, and advanced quantized computer-vision models for segmentation and action recognition. His work spans low-level inference engineering, operator correctness, benchmarking, and unit-test-driven enhancements—bridging research-grade models to production-ready, performance-tuned implementations. Notably, he has a knack for spotting and fixing subtle type and inference issues in large C++ codebases that materially improve model robustness and deployability.
7 years of coding experience
Github Skills (11)
neural-network10
quantization10
computer-vision10
mxnet10
quants10
deeplearning-ai10
c-language10
deep-learning10
cprogramming-language10
image-classification9
tensorflow3
Programming languages (4)
Github contributions (5)
dmlc/gluon-cv
Aug 2019 - Dec 2019
Gluon CV Toolkit
Role in this project:
ML Engineer
ML EngineerContributions:6 commits, 7 PRs, 1 push in 4 months
Contributions summary:Wuxun primarily contributed to the development and integration of quantized models within the GluonCV toolkit. Their work involved enabling and improving quantized versions of various computer vision models like FCN, PSPNet, and Deeplab, specifically for the task of image segmentation. They also added support for action recognition models, demonstrating a focus on model optimization and performance enhancements using quantization techniques, performance numbers, and unittests. Additionally, they refactored code to align with new mxnet standards.
pytorchmxnetperson-reidsemantic-segmentationdeep-learning
apache/mxnet
May 2019 - Apr 2020
Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
Role in this project:Back-end Developer
Back-end DeveloperContributions:15 commits, 21 PRs, 125 comments in 11 months
Contributions summary:Wuxun primarily contributed to the C++ codebase of the MXNet deep learning framework. Their work involved fixing type inconsistencies when loading quantized parameters, improving the C++ inference script to support benchmarks, and addressing issues related to the flatten operation and dropout masks. They also integrated MKL-DNN for Conv3d and Pool3d/1d operators and made improvements related to the BilinearResize2D operator. Their contributions span multiple parts of the codebase, including core operators and testing infrastructure.
pythonschedulerdataflowmutationdata-science
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.