Justin Kerr is a PhD candidate and AI researcher with a decade of engineering experience focused on high-performance machine learning systems. Based at Berkeley, he contributes to cutting-edge open-source NeRF tooling—optimizing NeRFstudio’s Gaussian Splatting pipeline by integrating PyTorch JIT, improving sampling schedules, and overhauling rendering and depth calculations to accelerate training and inference. He blends deep research insight with practical engineering, shipping performance-sensitive changes across rendering, parameter storage, and visualization components. Notably, his work reduces computational bottlenecks in real-world 3D reconstruction workflows, making advanced neural rendering more usable for practitioners.
Contributions:200 reviews, 23 commits, 157 PRs in 3 months
Contributions summary:Justin primarily contributed to optimizing the NeRFstudio project, focusing on improving the efficiency and performance of Gaussian Splatting models. Their work involved integrating PyTorch JIT compilation to speed up computations, and implementing a sampling schedule to accelerate training. The user also made significant changes to the rendering pipeline, including adding alpha rendering, fixing depth calculation bugs, and refactoring the gaussian parameter storage. They also integrated improvements to the viewer.
Contributions:3 commits, 2 pushes, 1 branch in 9 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.