This repository is a curated list of papers, resources, and implementations related to 3D Gaussian Splatting (3DGS). It serves as a comprehensive hub for researchers and developers interested in this rapidly evolving field, providing access to foundational concepts, cutting-edge advancements, and practical tools.

How It Works

The project focuses on 3D Gaussian Splatting, a technique that represents 3D scenes using millions of 3D Gaussians. This explicit representation allows for efficient, high-quality rendering at real-time speeds, overcoming limitations of previous implicit neural radiance field methods. The list covers various aspects, including papers on core algorithms, implementations across different frameworks (Taichi, PyTorch, C++/CUDA), viewers for game engines and web, supporting tools, and learning resources.

Quick Start & Requirements

This repository is a list of resources, not a runnable software package. To use specific implementations, refer to their respective project pages for installation and execution instructions. Dependencies vary widely, often including Python, CUDA, and specific deep learning frameworks.

Highlighted Details

• Comprehensive Coverage: Includes papers, implementations, viewers, tools, and learning resources for 3DGS.
• Diverse Implementations: Features various approaches like Taichi, Nerfstudio integration, high-performance C++/CUDA, and cross-platform solutions.
• Broad Application Areas: Covers use cases in autonomous driving, avatar creation, dynamic scenes, compression, and more.
• Active Curation: Regularly updated with new research and code releases, reflecting the fast-paced nature of the field.

Maintenance & Community

The list is actively maintained by the community, with contributions encouraged. Links to related projects (MrNeRF) and a contributing guide are provided.

Licensing & Compatibility

Implementations listed have various licenses, including Apache-2.0 and AGPL-3.0. Users must check individual licenses for compatibility with commercial or closed-source projects.

Limitations & Caveats

As a curated list, the repository itself does not have technical limitations. However, the rapid pace of research means that specific implementations may become outdated or superseded by newer methods. Users should verify the status and maintenance of individual linked projects.