What it does

This is a curated list—an “awesome list” in the classic GitHub tradition—cataloging papers, code implementations, viewers, tools, and learning resources for 3D Gaussian Splatting (3DGS). It covers the ecosystem from the original Inria reference implementation through community ports in Taichi, Warp, and C++/CUDA, plus game engine plugins for Unity and Unreal, web viewers in WebGL and WebGPU, and even ROS2 and OpenXR VR support.

The interesting bit

The list reveals how fast a research technique can colonize every runtime in existence. There are now at least three competing Unity plugins, two Unreal plugins, viewers for iOS Metal and visionOS, and format converters juggling PLY, SPZ, SPLAT, and KSPLAT. The project also maintains a searchable web database of papers, acknowledging that manual README maintenance can’t keep pace with the field.

Key highlights

• Searchable papers database at mrnerf.github.io, separate from the flat README list
• Implementations span Python/CUDA, C++, Taichi, Warp, and even a Jupyter notebook for 2D Gaussian Splatting
• Game engine support: Unity (3 plugins), Unreal (2 plugins), PlayCanvas
• Viewers cover WebGL, WebGPU, desktop OpenGL/Vulkan, iOS Metal, and OpenXR VR
• Tools include browser-based editors (SuperSplat), Houdini integration, and multiple format converters
• Learning resources from MIT coursework to HuggingFace blog posts and Metaverse Standards Forum talks

Caveats

• The README is a flat list with minimal curation or quality ranking; some entries are just links with no description
• No clear criteria for what qualifies as “awesome” versus merely existing
• Several sections appear duplicated or near-duplicated between the top and bottom of the file

Verdict

Essential if you’re trying to navigate the 3DGS ecosystem without drowning in arXiv notifications. Skip it if you already have a working pipeline and aren’t shopping for alternatives.