Changqian Yu
Director at Kling AI
Beijing, China
Summary
🤩
Rockstar🎓
Top SchoolChangqian Yu is a computer vision and multimodal foundation model leader with 11 years of experience, currently heading Kling Image at Kuaishou Technology from Beijing. He blends deep research training—a PhD in computer vision with joint work at Huazhong University and the University of Adelaide—with hands-on engineering roles at Meituan, Kunlun, Microsoft, and Megvii. His work spans image foundation models, semantic segmentation, and vision–language models, and he contributes practical tooling such as the TorchSeg PyTorch reference implementation used for fast, modular segmentation development. Known for moving research into production, he has steered core model architectures (ResNet/Xception) and dataset/engine integrations to accelerate training and evaluation pipelines. Colleagues value his mix of academic rigor and production-first pragmatism, and his background in autonomous-vehicle vision research hints at strong real-world system deployment experience.
11 years of coding experience
4 years of employment as a software developer
Joint Doctor of Philosophy - Joint PhD Computer Vision, Joint Doctor of Philosophy - Joint PhD Computer Vision at The University of Adelaide CollegeDoctor of Philosophy - PhD Computer Vision, Doctor of Philosophy - PhD Computer Vision at Huazhong University of Science and TechnologyBachelor's degree Automation, Bachelor's degree Automation at Shandong University
Github Skills (9)
semantic-segmentation10
pytorch10
xception10
resnet10
python10
image-processing9
machine-learning9
deep-learning9
computer-vision9
Programming languages (7)
Github contributions (5)
yu-changqian/TorchSeg
Jan 2019 - Mar 2020
Fast, modular reference implementation and easy training of Semantic Segmentation algorithms in PyTorch.
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:1 release, 83 commits, 1 PR in 1 year 2 months
Contributions summary:Changqian primarily focused on building and modifying base models for semantic segmentation. This included implementing ResNet and Xception architectures and integrating them into the project's framework. The user also made changes to common datasets and core engine components, and added and modified util functions for image processing, indicative of work that supports model training and evaluation for semantic segmentation tasks.
pytorchsemantic-segmentation
awfssv/pytorch-cn
Apr 2017 - Jun 2017
Pythrch-CN文档地址
Contributions:55 commits, 17 PRs, 1 comment in 2 months
python
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.