Jeffrey Hu

Hi, I'm Jeffrey! I'm an aspiring PhD student for Fall 2025 and an active contributor to gsplat.

Prevously, I worked on autonomous vehicles and building the perception system for autonomous trains at Parallel Systems. I received my undergraduate and Master's degrees in Electrical Engineering and Computer Science at MIT with Professor Antonio Torralba.

Email  /  CV  /  Scholar  /  X  /  Linkedin  /  Github

Research

I am deeply interested in 3D computer vision, computer graphics, and differentiable rendering. I want to capture and build generative models for the real world.

gsplat is an open-source library designed for training and developing Gaussian Splatting methods. It features a front-end with Python bindings compatible with the PyTorch library and a back-end with highly optimized CUDA kernels. gsplat offers numerous features that enhance the optimization of Gaussian Splatting models, which include optimization improvements for speed, memory, and convergence times. Experimental results demonstrate that gsplat achieves up to 10% less training time and 4x less memory than the original implementation. Utilized in several research projects, gsplat is actively maintained on GitHub. Source code is available at this https URL under Apache License 2.0. We welcome contributions from the open-source community.

Contributions to gsplat

Improved densification strategy for 3DGS.

Use bilateral grids to model changes in camera exposure and other ISP effects for gaussian splats.

Use image compression to compress gaussian splats.

Rewrite 3DGS projection function to support fisheye cameras.

Interpret the z-axis of 3D Gaussians as their normal vector to achieve a similar effect to 2DGS.

Implement camera defocus blur modeling in gsplat.

Miscellanea

The template is stolen from Jon Barron.