MinĀ Chong
Research Assistant at University of Illinois at Urbana-Champaign
Champaign, Illinois, United States
Summary
š¤©
Rockstarš
Top SchoolMin Chong is a computer vision researcher and PhD candidate at UIUC with nine years of experience building and deploying generative models, image-to-image translation, and adversarial-robust systems. As a long-term research assistant and former intern at Google, Adobe, ByteDance, and Revery.AI, Min has driven projects from unsupervised semantic editing of StyleGAN outputs to accelerating GAN training with pretrained networks. Their open-source work includes practical contributions to high-profile repos like JoJoGAN and GANsNRoses, focusing on face stylization, projection/alignment, and making inference accessible on CPU/Colab environments. Min combines rigorous academic training and production-minded engineering, often bridging research prototypes into usable tools for the community. Unusually for an ML researcher, Min also brings leadership and technical resilience forged during military service as a combat engineer. Based in Champaign, Illinois, they continuously explore disentanglement and robustness techniques that improve both model interpretability and real-world reliability.
9 years of coding experience
2 years of employment as a software developer
University of Illinois Urbana-Champaign
English, Chinese
Github Skills (9)
computer-vision10
pytorch10
machine-learning10
image-translation10
python10
animate9
inference9
google-colaboratory7
google-colab7
Programming languages (4)
Github contributions (5)
mchong6/GANsNRoses
Jun 2021 - May 2022
Official PyTorch repo for GAN's N' Roses. Diverse im2im and vid2vid selfie to anime translation.
Role in this project:
ML Engineer
ML EngineerContributions:34 commits, 3 PRs, 23 pushes in 11 months
Contributions summary:Min's commits primarily involve modifications to inference-related code within a PyTorch repository focused on image translation. They removed CUDA-related code, indicating a focus on adapting the inference process for environments without GPU support like Colab. The user also added and updated a Colab notebook, showing a direct contribution to the project's usability and accessibility for users.
pytorchdiversedeep-learningtranslationselfie
mchong6/JoJoGAN
Dec 2021 - Feb 2022
Official PyTorch repo for JoJoGAN: One Shot Face Stylization
Role in this project:ML Engineer
ML EngineerContributions:74 commits, 5 PRs, 69 pushes in 1 month
Contributions summary:Min primarily worked on modifications related to the `e4e` model, indicating a focus on face stylization and image translation. The user's contributions include incorporating and integrating the `e4e` model, as well as adjusting model parameters and code within the model files. The user also implemented and refined face alignment and projection functionalities for use within the style transfer pipeline. Overall, the commits are focused on the image transformation aspect of the JoJoGAN project.
pytorchone-shotdeep-learningcomputer-visiongans
Find and Hire Top DevelopersWeāve 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.