# FastGS **Repository Path**: Steven-wei/FastGS ## Basic Information - **Project Name**: FastGS - **Description**: No description available - **Primary Language**: Unknown - **License**: MIT - **Default Branch**: main - **Homepage**: None - **GVP Project**: No ## Statistics - **Stars**: 0 - **Forks**: 0 - **Created**: 2026-02-02 - **Last Updated**: 2026-02-02 ## Categories & Tags **Categories**: Uncategorized **Tags**: None ## README

FastGS: Training 3D Gaussian Splatting in 100 Seconds

[🌐 Homepage](https://fastgs.github.io/) | [πŸ“„ Paper](https://arxiv.org/abs/2511.04283) ｜[πŸ€— Pre-trained model](https://huggingface.co/Goodsleepeverday/fastgs)

## πŸš€ What Makes FastGS Special? FastGS is a **general acceleration framework** that supercharges 3D Gaussian Splatting training while maintaining Comparable rendering quality. Our method stands out with: - **⚑ Blazing Fast Training**: Achieve SOTA results within **100 seconds**. **3.32Γ— faster** than DashGaussian on Mip-NeRF 360 dataset. **15.45Γ— acceleration** vs vanilla 3DGS on Deep Blending. - **⚑ High fidelity**: Comparable rendering quality with SOTA methods - **🎯 Easy Integration**: Seamlessly integrates with various backbones (Vanilla 3DGS, Scaffold-GS, Mip-splatting, etc.) - **πŸ› οΈ Multi-Task Ready**: Proven effective across dynamic scenes, surface reconstruction, sparse-view, large-scale, and SLAM tasks - **πŸ’‘ Memory-Efficient**: Low GPU Memory requirements make it accessible for various hardware setups - **πŸ”§ Easy Deployment**: Simple post-training tool for feedforward 3DGS that works out-of-the-box ## πŸ“’ Latest Updates #### πŸ”₯ **[2025.11.16]** Code Released - Get Started Now! πŸš€ #### πŸ”₯ **[2025.11.17]** Pre-trained model Released πŸ€—! #### πŸ“„ **[2025.11.26]** The supplementary material has been released [here](https://arxiv.org/abs/2511.04283)! #### πŸ”§ **[2025.11.27]** The tutorial has been released β€” see the [Wiki](https://github.com/fastgs/FastGS/wiki)! #### πŸ”₯ **[2025.11.29]** The dynamic scene reconstruction code [Fast-D3DGS](https://github.com/fastgs/FastGS/tree/fast-d3dgs) has been released! #### πŸ”₯ **[2025.12.03]** The sparse-view reconstruction code [Fast-DropGaussian](https://github.com/fastgs/FastGS/tree/fast-dropgaussian) has been released! #### πŸ₯‡ **[2026.01]** Our method was used as a component in the [winning solution](https://arxiv.org/pdf/2601.19489) (1st placeπŸ₯‡) of the **[SIGGRAPH Asia 2025 3DGS Fast Reconstruction Challenge](https://gaplab.cuhk.edu.cn/projects/gsRaceSIGA2025/index.html#awards)**. We sincerely thank the **3DV-CASIA** for their interest and adoption of our work. ## 🎯 Coming Soon #### Released Modules - **Dynamic Scenes Reconstruction** β€” [Fast-D3DGS](https://github.com/fastgs/FastGS/tree/fast-d3dgs) (based on [Deformable-3D-Gaussians](https://github.com/ingra14m/Deformable-3D-Gaussians)) β€” Released - **Sparse-view Reconstruction** β€” [Fast-DropGaussian](https://github.com/fastgs/FastGS/tree/fast-dropgaussian) (based on [DropGaussian](https://github.com/DCVL-3D/DropGaussian_release)) β€” Released #### To Be Released After Paper Acceptance - **Autonomous Driving Scenes** β€” [street_gaussians](https://github.com/zju3dv/street_gaussians) - **Surface Reconstruction** β€” [PGSR](https://github.com/zju3dv/PGSR) - **Large-scale Reconstruction** β€” [OctreeGS](https://github.com/city-super/Octree-GS/tree/main) - **SLAM** β€” [Photo-SLAM](https://github.com/HuajianUP/Photo-SLAM) - **Backbone Enhancing** β€” [Mip-splatting](https://github.com/autonomousvision/mip-splatting) ## πŸ—οΈ Training Framework Our training pipeline leverages **PyTorch** and optimized **CUDA extensions** to efficiently produce high-quality trained models in record time. ### πŸ’» Hardware Requirements - **GPU**: CUDA-ready GPU with Compute Capability 7.0+ - **Memory**: 24 GB VRAM (for paper-quality results; we recommend NVIDIA RTX4090) ### πŸ“¦ Software Requirements - **Conda** (recommended for streamlined setup) - **C++ Compiler** compatible with PyTorch extensions - **CUDA SDK 11** (or compatible version) - **⚠️ Important**: Ensure C++ Compiler and CUDA SDK versions are compatible ### ⚠️ CUDA Version Reference Our testing environment uses the following CUDA configuration: | Component | Version | |---------------------------------------|------------------| | Conda environment CUDA version | 11.6 | | Ubuntu system `nvidia-smi` CUDA | 12.2 | | `nvcc -V` compiler version | 11.8 (v11.8.89) | > **Note**: The Conda CUDA and system CUDA versions may differ. The compiler version (`nvcc`) is what matters for PyTorch extensions compilation (diff-gaussian-rasterization_fastgs). ## πŸš€ Quick Start ### πŸ“₯ Clone the Repository ```bash git clone https://github.com/fastgs/FastGS.git --recursive cd FastGS ``` ### βš™οΈ Environment Setup We provide a streamlined setup using Conda: ```shell # Windows only SET DISTUTILS_USE_SDK=1 # Create and activate environment conda env create --file environment.yml conda activate fastgs ``` ### πŸ“‚ Dataset Organization Organize your datasets in the following structure: ```bash datasets/ β”œβ”€β”€ mipnerf360/ β”‚ β”œβ”€β”€ bicycle/ β”‚ β”œβ”€β”€ flowers/ β”‚ └── ... β”œβ”€β”€ db/ β”‚ β”œβ”€β”€ playroom/ β”‚ └── ... └── tanksandtemples/ β”œβ”€β”€ truck/ └── ... ``` The MipNeRF360 scenes are hosted by the paper authors [here](https://jonbarron.info/mipnerf360/). You can find our SfM data sets for Tanks&Temples and Deep Blending [here](https://repo-sam.inria.fr/fungraph/3d-gaussian-splatting/datasets/input/tandt_db.zip). ## 🎯 Training & Evaluation ### ⚑ FastGS (Standard) Train the base model with optimal speed and quality balance: ```bash bash train_base.sh ``` ### 🎨 FastGS-Big (High Quality) For enhanced quality with slightly longer training time: ```bash bash train_big.sh ```
πŸ“‹ Advanced: Command Line Arguments for train.py #### --loss_thresh Threshold of the loss map; a lower value generally results in more Gaussians being retained. #### --grad_abs_thresh Absolute gradient (same as Abs-GS) threshold for split. #### --grad_thresh Gradient(same as vanilla 3DGS) threshold for clone. #### --highfeature_lr Learning rate for high-order SH coefficients (features_rest). #### --lowfeature_lr Learning rate for low-order SH coefficients (features_dc). #### --dense Percentage of scene extent (0--1) a point must exceed to be forcibly densified. #### --mult Multiplier for the compact box to control the tile number of each splat #### --source_path / -s Path to the source directory containing a COLMAP or Synthetic NeRF data set. #### --model_path / -m Path where the trained model should be stored (```output/``` by default). #### --images / -i Alternative subdirectory for COLMAP images (```images``` by default). #### --eval Add this flag to use a MipNeRF360-style training/test split for evaluation. #### --resolution / -r Specifies resolution of the loaded images before training. If provided ```1, 2, 4``` or ```8```, uses original, 1/2, 1/4 or 1/8 resolution, respectively. For all other values, rescales the width to the given number while maintaining image aspect. **If not set and input image width exceeds 1.6K pixels, inputs are automatically rescaled to this target.** #### --data_device Specifies where to put the source image data, ```cuda``` by default, recommended to use ```cpu``` if training on large/high-resolution dataset, will reduce VRAM consumption, but slightly slow down training. Thanks to [HrsPythonix](https://github.com/HrsPythonix). #### --white_background / -w Add this flag to use white background instead of black (default), e.g., for evaluation of NeRF Synthetic dataset. #### --sh_degree Order of spherical harmonics to be used (no larger than 3). ```3``` by default. #### --convert_SHs_python Flag to make pipeline compute forward and backward of SHs with PyTorch instead of ours. #### --convert_cov3D_python Flag to make pipeline compute forward and backward of the 3D covariance with PyTorch instead of ours. #### --debug Enables debug mode if you experience erros. If the rasterizer fails, a ```dump``` file is created that you may forward to us in an issue so we can take a look. #### --debug_from Debugging is **slow**. You may specify an iteration (starting from 0) after which the above debugging becomes active. #### --iterations Number of total iterations to train for, ```30_000``` by default. #### --ip IP to start GUI server on, ```127.0.0.1``` by default. #### --port Port to use for GUI server, ```6009``` by default. #### --test_iterations Space-separated iterations at which the training script computes L1 and PSNR over test set, ```7000 30000``` by default. #### --save_iterations Space-separated iterations at which the training script saves the Gaussian model, ```7000 30000 ``` by default. #### --checkpoint_iterations Space-separated iterations at which to store a checkpoint for continuing later, saved in the model directory. #### --start_checkpoint Path to a saved checkpoint to continue training from. #### --quiet Flag to omit any text written to standard out pipe. #### --feature_lr Spherical harmonics features learning rate, ```0.0025``` by default. #### --opacity_lr Opacity learning rate, ```0.05``` by default. #### --scaling_lr Scaling learning rate, ```0.005``` by default. #### --rotation_lr Rotation learning rate, ```0.001``` by default. #### --position_lr_max_steps Number of steps (from 0) where position learning rate goes from ```initial``` to ```final```. ```30_000``` by default. #### --position_lr_init Initial 3D position learning rate, ```0.00016``` by default. #### --position_lr_final Final 3D position learning rate, ```0.0000016``` by default. #### --position_lr_delay_mult Position learning rate multiplier (cf. Plenoxels), ```0.01``` by default. #### --densify_from_iter Iteration where densification starts, ```500``` by default. #### --densify_until_iter Iteration where densification stops, ```15_000``` by default. #### --densify_grad_threshold Limit that decides if points should be densified based on 2D position gradient, ```0.0002``` by default. #### --densification_interval How frequently to densify, ```100``` (every 100 iterations) by default. #### --opacity_reset_interval How frequently to reset opacity, ```3_000``` by default. #### --lambda_dssim Influence of SSIM on total loss from 0 to 1, ```0.2``` by default. #### --percent_dense Percentage of scene extent (0--1) a point must exceed to be forcibly densified, ```0.01``` by default.

Note that similar to MipNeRF360 and vanilla 3DGS, we target images at resolutions in the 1-1.6K pixel range. For convenience, arbitrary-size inputs can be passed and will be automatically resized if their width exceeds 1600 pixels. We recommend to keep this behavior, but you may force training to use your higher-resolution images by setting ```-r 1```. ## 🎬 Interactive Viewers Our 3DGS representation is identical to vanilla 3DGS, so you can use the official [SIBR viewer](https://github.com/graphdeco-inria/gaussian-splatting?tab=readme-ov-file#interactive-viewers) for interactive visualization. For a quick start without local setup, try the web-based [Supersplat](https://superspl.at/editor). ## 🎯 Quick Facts | Feature | FastGS | Previous Methods | |---------|---------|---------------------| | Training Time | **100 seconds** | 5-30 minutes | | Gaussian Efficiency | βœ… **Strict Control** | ❌ Redundant Growth | | Memory Usage | βœ… **Low Footprint** | ❌ High Demand | | Task Versatility | βœ… **6 Domains** | ❌ Limited Scope | ## πŸ“§ Contact If you have any questions, please contact us at **renshiwei@mail.nankai.edu.cn**. ## πŸ™ Acknowledgements This project is built upon [3DGS](https://github.com/graphdeco-inria/gaussian-splatting), [Taming-3DGS](https://github.com/humansensinglab/taming-3dgs), [Speedy-Splat](https://github.com/j-alex-hanson/speedy-splat), and [Abs-GS](https://github.com/TY424/AbsGS). We extend our gratitude to all the authors for their outstanding contributions and excellent repositories! **License**: Please adhere to the licenses of 3DGS, Taming-3DGS, and Speedy-Splat. Special thanks to the authors of [DashGaussian](https://github.com/YouyuChen0207/DashGaussian) for their generous support! ## Citation If you find this repo useful, please cite: ``` @article{ren2025fastgs, title={FastGS: Training 3D Gaussian Splatting in 100 Seconds}, author={Ren, Shiwei and Wen, Tianci and Fang, Yongchun and Lu, Biao}, journal={arXiv preprint arXiv:2511.04283}, year={2025} } ``` ---
**⭐ If FastGS helps your research, please consider starring this repository!** *FastGS: Training 3D Gaussian Splatting in 100 Seconds*
---