RS-AC-3DGS

Overview

RS-AC-3DGS provides tools to process and convert SLAM results from Active Camera to 3D Gaussian Splatting format. The toolkit then enables training using popular 3D Gaussian Splatting methods and visualization of the results.

Installation

You can either use this repository or directly clone the official 3D Gaussian Splatting repository:

# Option 1: Use this repository
git clone https://github.com/RoboSense-Robotics/robosense_ac_3dgs.git

# Option 2: Use official repository
git clone https://github.com/graphdeco-inria/gaussian-splatting.git

For detailed installation instructions, please refer to the 3D Gaussian Splatting README.

Usage

1. Data Preprocessing

Follow the instructions in preprocess/README.md to convert SLAM data to 3DGS format.

2. Training

cd gaussian-splatting
python train.py -s ../preprocess/data/[dataset_name]/scene

3. Visualization

For visualization setup, refer to the Interactive Viewers section in the original repository.

The gaussian-splatting/SIBR_viewers in this repository uses the fossa_compatibility branch. If you need to use other branches, you can download and install them from the official repository (https://gitlab.inria.fr/sibr/sibr_core.git).

cd gaussian-splatting/SIBR_viewers/install/bin
./SIBR_gaussianViewer_app -m /path/to/output/3D-Gaussian-Splatting/[exp_name] --rendering-size 1920 1080

Additional Visualization Tools

1. Video Recording:

   • Follow the tutorial at YouTube Guide

2. Online Viewer:

   • SuperSplat Editor - Alternative web-based visualization tool

License

This project is licensed under the MIT License - see the LICENSE file for details.

Acknowledgments

• 3D Gaussian Splatting

Contact

For questions and support, please open an issue on our GitHub repository.