Baike.dev
All toolsOnline toolsAI codingTrendingOpen sourceNewsSubmit
Log in
< Back to tools
C

c3dgs

> 编程语言
Open source

Compressed 3D Gaussian Splatting for Accelerated Novel View Synthesis

400 stars0 likes1 views
WebsiteGitHub

About

Compressed 3D Gaussian Splatting for Accelerated Novel View Synthesis

Abstract

Recently, high-fidelity scene reconstruction with an optimized 3D Gaussian splat representation has been introduced for novel view synthesis from sparse image sets. Making such representations suitable for applications like network streaming and rendering on low-power devices requires significantly reduced memory consumption as well as improved rendering efficiency. We propose a compressed 3D Gaussian splat representation that utilizes sensitivity-aware vector clustering with quantization-aware training to compress directional colors and Gaussian parameters. The learned codebooks have low bitrates and achieve a compression rate of up to 31x on real-world scenes with only minimal degradation of visual quality. We demonstrate that the compressed splat representation can be efficiently rendered with hardware rasterization on lightweight GPUs at up to 4x higher framerates than reported via an optimized GPU compute pipeline. Extensive experiments across multiple datasets demonstrate the robustness and rendering speed of the proposed approach.

Citation

If you find our work useful, please cite:

@misc{niedermayr2023compressed,
    title={Compressed 3D Gaussian Splatting for Accelerated Novel View Synthesis}, 
    author={Simon Niedermayr and Josef Stumpfegger and Rüdiger Westermann},
    year={2023},
    eprint={2401.02436},
    archivePrefix={arXiv},
    primaryClass={cs.CV}
}

Installation

Requirements

  • Conda (recommended for easy setup)
  • C++ Compiler for PyTorch extensions
  • CUDA SDK 12 for PyTorch extensions
  • C++ Compiler and CUDA SDK must be compatible

Please refer to the original 3D Gaussian Splatting repository for more details about requirements.

Cloning the Repository

bash
git clone https://github.com/KeKsBoTer/c3dgs

Setup

Our default, provided install method is based on Conda package and environment management:

SET DISTUTILS_USE_SDK=1 # Windows only
conda env create --file environment.yml
conda activate c3dgs

Running

To compress a scene reconstructed by 3D Gaussian Splatting, simply use:

bash
python compress.py --model_path  --data_device "cuda" --output_vq 

Note: After the compression is complete the script will compute the metrics (PSNR, SSIM, LPIPS) for the test images.

Evaluation

For a more detailed evaluation and rendering run:

bash
python render.py -m  # Generate renderings
python metrics.py -m  # Compute error metrics on renderings

Interactive Viewers

Our renderer supports direct rendering of the compressed files on many platforms with improved rendering speed. It can be found here.

Alternatively, you can convert the compressed .npz scene files back to .ply files and open them with the SIBR Viewer:

bash
python npz2ply.py  [--ply_file ]

Issues· 11 open

View all issuesOpen on GitHub

No open issues yet, or sync has not completed.

> Tags

Python3dgscompressiongaussiannerf

No comments yet. Be the first to share.

> Details

PublishedAug 1, 2026
UpdatedSep 17, 2026
Category编程语言
PricingOpen source

> Related tools

T
TypeScript
JavaScript 的超集,为前端与全栈提供静态类型
P
Python
通用编程语言,广泛用于 Web、数据与 AI
G
Go
Google 推出的简洁高效系统语言