Guilin Liu
Principal Research Scientist & Team Leader at NVIDIA
San Jose, California, United States
Summary
🤩
Rockstar🎓
Top SchoolGuilin Liu is a senior (staff-level) research scientist at NVIDIA with a decade of experience specializing in deep learning for graphics, vision, and generative models. Since joining NVIDIA in 2017 he has contributed to flagship technologies like DLSS and presented research demos alongside Jensen Huang, with two research videos exceeding one million views. His work spans model architecture, padding schemes, and loss function refinement—evidenced by contributions to NVIDIA’s partialconv repositories improving partial convolution layers and masking logic. Guilin holds a PhD in Computer Science from George Mason University and brings prior research internships at Adobe and TTIC plus early work on 3D point cloud processing for disaster evaluation. He blends rigorous academic training with production-grade engineering, delivering research that transitions into widely used GPU features embraced by millions of gamers. Notably, his contributions reveal a knack for turning advanced research ideas into highly optimized, deployable systems.
11 years of coding experience
4 years of employment as a software developer
Bachelor's degree International School of Software (major), Finance (minor), Bachelor's degree International School of Software (major), Finance (minor) at Wuhan UniversityDoctor of Philosophy (PhD) Computer Science, Doctor of Philosophy (PhD) Computer Science at George Mason University
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (7)
neural-network10
computer-vision10
machine-learning10
pytorch10
convolutional-neural-networks10
python9
loss-functions9
Programming languages (2)
Github contributions (5)
NVIDIA/partialconv
Nov 2018 - Jun 2020
A New Padding Scheme: Partial Convolution based Padding
Role in this project:
ML Engineer
ML EngineerContributions:17 commits, 2 PRs, 33 pushes in 1 year 6 months
Contributions summary:Guilin primarily focused on updating and modifying core model components related to partial convolutions. They refined the `partialconv2d.py` and `partialconv3d.py` files, modifying the convolution layers and masking logic. Additionally, the user made alterations to the `loss.py` file, suggesting involvement in model training or evaluation and the associated loss functions, likely for the partial convolution architecture.
convolution
liuguilin1225/liuguilin1225.github.io
Jun 2018 - Jul 2026
Contributions:152 pushes, 3 branches in 8 years 2 months
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.