awesome-3D-gaussian-splatting

Awesome 3D Gaussian Splatting

Browse the Paper List | LichtFeld Studio | Contribute | MrNeRF

Contents

• Papers & Documentation
• Implementations
• Viewers & Game Engine Support
• Tools & Utilities
• Learning Resources
• Credits

Papers & Documentation

Papers Database

Visit our comprehensive, searchable database of 3D Gaussian Splatting papers: Papers Database

Courses & Tutorials

• MIT Inverse Rendering Lectures (Module 2) - Academic deep dive into inverse rendering
• 3DGS Tutorial - Tutorial from the authors of the original 3DGS paper

Datasets

• NERDS 360 Multi-View dataset - High-quality outdoor scene dataset

Implementations

Official Reference

• Original Gaussian Splatting - The reference implementation by the original authors

Community Implementations

Implementation	Language	License	Description
LichtFeld Studio (lichtfeld.io)	C++/CUDA	GPL-3.0	The modular workstation for 3D Gaussian Splatting — train, inspect, edit, automate, and export from a single native app
Nerfstudio gsplat	Python/CUDA	Apache-2.0	Integration with Nerfstudio
OpenSplat	C++/CPU/GPU	AGPL-3.0	Cross-platform solution
Taichi 3D GS	Taichi	Apache-2.0	Taichi-based implementation
taichi-splatting	Taichi/PyTorch	Apache-2.0	Modular rasterizer for Taichi and PyTorch
Grendel Distributed 3DGS	Python/CUDA	Apache-2.0	Multi-GPU distributed training
Warp 3DGS	Warp/Python	AGPL-3.0	Warp-based implementation
RI3D	Python/CUDA	Unlicense	Few-shot gaussian splatting pipeline
gaussian_splatting	Python/CUDA	MIT	Readable implementation with a written derivation of the math
3d-gaussian-splatting	Python/CUDA	MIT	Compact reimplementation
gaussian_splatting_3d	Python/CUDA		Early community reimplementation
My-exp-Gaussians	Python/CUDA		Enhances the ability of 3D Gaussians to model complex scenes
360-gaussian-splatting	Python		Trains splats directly from 360° images
2D Gaussian Splatting	Jupyter	MIT	Notebook walkthrough of 2D gaussian splatting
DGSO	Python	MIT	Style transfer applied during gaussian optimization
3D-Gaussian-Splatting-Reconstruction	Python	Apache-2.0	Pure PyTorch implementation from video input — no CUDA compilation required, supports CPU/NVIDIA GPU backends, built-in pose estimator & GUI
MetalSplat	PyTorch/Metal	MIT	Metal-based implementation for Apple Silicon GPUs on PyTorch with automatic kernel compilation at runtime

Frameworks

• Pointrix - Differentiable point-based rendering
• msplat - Modular differential gaussian rasterization library
• GauStudio - Unified framework with multiple implementations
• DriveStudio - Urban scene reconstruction framework
• GSCodecStudio - Compression and Dynamic splattings
• gaussian-splatting-lightning - Derived algorithms plus an interactive web viewer

Viewers & Game Engine Support

Game Engines

• Unity Plugin
• Unity Plugin (gsplat-unity)
• Unity Plugin (DynGsplat-unity) - For dynamic splattings
• Unreal Plugin (MLSLabsGaussianSplattingRenderer-UE)
• Unreal Plugin (XScene-UEPlugin)
• PlayCanvas Engine
• Godot Plugin (gdgs) - Real-time 3DGS rendering plugin for Godot 4.3+

Web Viewers

WebGL

• Splat Viewer
• Gauzilla
• Interactive Viewer
• GaussianSplats3D
• gsplat.js
• A-Frame
• splaTV - Viewer for 4D Gaussians, with a live demo
• WebRTC viewer

WebGPU

• EPFL Viewer
• WebGPU Splat
• gaussian-splatting-webgpu
• SuperSplat Viewer - High-performance splat viewer
• PlayCanvas Model Viewer - Viewer for glTF and 3DGS assets

Desktop Viewers

• 3DGS.cpp - C++/Vulkan renderer for Windows, macOS, Linux, iOS and visionOS
• vkgs - Cross-platform C++/Vulkan renderer
• splatviz - Edit the rendering code at runtime or display multiple scenes at once
• OpenGL Viewer - PyOpenGL viewer, also with official CUDA backend
• Taichi Viewer - Renderer with benchmarking capability
• DearGaussianGUI
• LiteViz-GS
• Nerfstudio Viser
• Nerfstudio (gaussian_splatting branch)
• Jupyter notebook viewer

Native Applications

• Blender Add-on
• Blender Add-on (KIRI)
• Blender Add-on (404—GEN)
• Houdini Viewport Renderer - HDK/GLSL implementation of Gaussian Splatting in Houdini
• iOS Metal Viewer
• VR Support (OpenXR)
• ROS2 Support
• Splat Local - Video-to-3DGS pipeline running fully locally on Apple Silicon (COLMAP/GLOMAP poses, Metal-native training via Brush), with training checkpoints streamed live into a browser viewer

Tools & Utilities

Data Processing

• Kapture - Unified data format for visual localization
• Kapture image cropper - Undistorted image cropper to remove black borders
• 3DGS Converter - Format conversion tool
• Point Cloud Editor - Web-based point cloud editing
• SPZ Converter - SPZ conversion tool
• gsbox Converter - PLY SPLAT SPZ SPX conversion tool
• SplatTransform - CLI tool and Node/browser library for converting and editing splats, reads PLY, SOG, SPZ, SPLAT, KSPLAT and LCC/LCC2, writes PLY, SOG, SPZ, GLB, CSV, LOD and WebP
• GaussForge - C++/WASM-based conversion between PLY, SPZ, SPLAT, and KSPLAT
• SpectacularAI - Conversion scripts for different 3DGS conventions
• VGGT Factor Refinement - COLMAP-free pipeline using VGGT + factor graph, from video to COLMAP-format output
• splatreg - pip-installable splat registration: align & merge two 3DGS scans into one SE(3)/Sim(3) frame (recovers scale), CLI + pure-PyTorch API, no manual gizmo
• AURA - Calibrated per-splat confidence for 3DGS assets: held-out reliability labels, isotonic calibration, and a distribution-free conformal pruning certificate with a certified LOD ladder; exports via glTF/OpenUSD/SPZ (pip install aura-splat)
• Open Reality - Phone video to a 3D scene (VGGT-SLAM based, splat export with optional gsplat refinement) that an AI assistant can query over MCP: measurements, floor and wall planes, path planning, object list, robot-training exports; self-hostable

Development Tools

• GSOPs for Houdini - Houdini integration tools
• camorph - Camera parameter conversion
• SuperSplat - Free, open source browser-based 3DGS editor with one-click publishing

Learning Resources

Blog Posts

• 3DGS Introduction - HuggingFace guide
• Comprehensive overview of Gaussian Splatting
• Very good (technical) intro to 3D Gaussian Splatting
• Gaussian Splatting is pretty cool
• Making Gaussian Splats smaller
• Making Gaussian Splats more smaller
• Compressing Gaussian Splats
• PlayCanvas Open Sources SOG - Spatially Ordered Gaussians, a super-compressed format based on WebP that is 15-20x smaller than PLY
• Implementation Details - Technical deep dive
• Mathematical Foundation - Theory explanation
• Mathematical details of forward and backward passes
• PyTorch Implementation - Curated implementation of Vanilla 3DGS in PyTorch
• NeRFs vs. 3DGS
• Gaussian Head Avatars: A Summary
• 3D in Geospatial: NeRFs, Gaussian Splatting, and Spatial Computing
• Capture Guide - Image capture tutorial
• Discussion about gs universal format

Talks

• Gaussian Splats: Ready for Standardization? - Metaverse Standards Forum 1/28/2025
• Unity Integration Guide - Metaverse Standards Forum 5/6/2025

Video Tutorials

• Getting Started (Windows)
• Two-Minute Explanation
• Computerphile 3DGS explanation
• Gaussian Splats Town Hall - Part 2
• Intro to gaussian splatting (and Unity plugin)
• Jupyter Tutorial

YouTube Channels

• LichtFeld Studio - Tutorials, release walkthroughs and dev updates for LichtFeld Studio

Credits

• Thanks to Leonid Keselman for informing me about the release of the paper "Real-time Photorealistic Dynamic Scene Representation and Rendering with 4D Gaussian Splatting".
• Thanks to Eric Haines for suggesting the jupyter notebook viewer, windows tutorial and for fixing text hyphenations and other issues.
• Thanks to Henry Pearce for maintaining contributions.
• Yehe Liu