GSCodec Studio is an open-source framework for Gaussian Splats Compression, including static and dynamic splats representation, reconstruction and compression. It is bulit upon an open-source 3D Gaussian Splatting library gsplat, and extended to support 1) dynamic splats representation, 2) training-time compression simulation, 3) more test-time compression strategies.
# Clone the repo.
git clone https://github.com/JasonLSC/GSCodec_Studio.git --recursive
cd GSCodec_Studio
# Make a conda environment
conda create --name gscodec_studio python=3.12
conda activate gscodec_studioPlease install Pytorch first.
Then, you can install the gsplat library extended with more compression features from source code. In this way it will build the CUDA code during installation.
pip install .
# pip install -e . (develop mode)Preparations
Same as gsplat, we need to install some extra dependencies and download the relevant datasets before the evaluation.
cd examples
pip install -r requirements.txt
# download mipnerf_360 benchmark data
python datasets/download_dataset.py
# or place other dataset under 'data' folder
ln -s data/tandt /xxxx/Dataset/tandtWe also use third-party library, 'python-fpnge', to accelerate image saving operations during the experiment for now. We also use third-party library, 'gridencoder', to facilitate hash encoding.
cd ..
pip install third_party/python-fpnge-master
pip install third_party/gridencoderBefore we start running scripts, we also need to install library for vector quantization and plas sorting.
# refer to https://github.com/DeMoriarty/TorchPQ?tab=readme-ov-file#install to see how to install TorchPQ
pip install git+https://github.com/fraunhoferhhi/PLAS.gitStatic Gaussian Splats Training and Compression
We provide a script that enables more memory-efficient Gaussian splats while maintaining high visual quality, such as representing the Truck scene with only about 8MB of storage. The script includes 1) the static splats training with compression simulation, 2) the compression of trained static splats, and 3) the metric evaluation of uncompressed and compressed static splats.
# Tanks and Temples dataset
bash benchmarks/compression/final_exp/mcmc_tt_sim.shDynamic Gaussian Splats Training and Compression
First, please follow the dataset preprocessing instruction described in the file for training data prepration.
Next, run the script for dynamic gaussian splats training and compression.
cd examples
bash benchmarks/dyngs/dyngs.shExtract Per-Frame Static Gaussian from Dynamic Splats
If you finsh the training of dynamic splats, then you can use the script to extract static gaussian splats stored at discrte timesteps in ".ply" file
cd examples
bash benchmarks/dyngs/export_plys.shLoad .ply File and Render
If you want to directly read and render splats exported as PLY files, you can use the following script. Note: You need to modify the PLY file path in the scripts.
cd examples
bash benchmarks/load_ply_and_render.shThis project is developed by the following contributors:
- Sicheng Li: [email protected]
- Chengzhen Wu: [email protected]
If you have any questions about this project, please feel free to contact us.
This project is bulit on gsplat. We thank all contributors from gsplat for building such a great open-source project.