Zhiqiang Wang
Deep Learning Algorithm Engineer at 7invensun
Chaoyang District, Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolZhiqiang Wang is a deep learning algorithm engineer with 10 years of experience specializing in computer vision and image processing, currently based in Chaoyang District, Beijing. He has applied his computational mathematics background to build and optimize detection pipelines, contributing notable improvements and tests to the widely used pytorch/vision repository. Zhiqiang also maintains practical deployment expertise—integrating and accelerating YOLOv5 across runtimes like TensorRT, ONNX, and TVM to meet real-world inference constraints. His career spans startups and product teams where he turned research-level models into production-ready systems, often simplifying test setups and refactoring type annotations for long-term maintainability. Colleagues appreciate that he blends rigorous numerical thinking with hands-on engineering to bridge algorithms and efficient deployment.
10 years of coding experience
2 years of employment as a software developer
Master of Science, Computational and Applied Mathematics, Master of Science, Computational and Applied Mathematics at Capital Normal UniversityBachelor's Degree, Mathematics and Applied Mathematics, Bachelor's Degree, Mathematics and Applied Mathematics at Ningbo University
Chinese, English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (13)
tensorrt10
object-detection10
machine-learning10
pytorch10
computer-vision10
inference10
deep-learning10
onnx10
optimisation10
optimization10
testing9
python9
cuda6
Programming languages (14)
Github contributions (5)
zhiqwang/yolort
Aug 2020 - Jan 2023
yolort is a runtime stack for yolov5 on specialized accelerators such as tensorrt, libtorch, onnxruntime, tvm and ncnn.
Role in this project:
ML Engineer
ML EngineerContributions:16 releases, 131 reviews, 447 commits in 2 years 5 months
Contributions summary:Zhiqiang primarily worked on integrating the yolov5 models into a runtime stack, optimizing them for deployment on specialized hardware accelerators. Their contributions focused on fixing build processes, enhancing compatibility, and improving the performance of the models for inference. They also addressed type annotation issues and refactored parts of the codebase, and tested various techniques such as ONNX and TensorRT to optimize the process.
ncnnruntimegraghsurgeontensorflowopenmmlab
pytorch/vision
Jul 2019 - May 2022
Datasets, Transforms and Models specific to Computer Vision
Role in this project:ML Engineer
ML EngineerContributions:65 reviews, 22 commits, 28 PRs in 2 years 10 months
Contributions summary:Zhiqiang primarily contributed to the `pytorch/vision` repository by implementing and testing computer vision algorithms, specifically related to anchor generators and object detection. They added tests with ground-truth outputs, simplified the setup for the anchor generator in unit tests, and refactored the default boxes calculations. The user also made documentation updates for torchvision ops, as well as replacing annotations with typing.
pytorchvisiondeep-learningdatasetcomputer-vision
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.