1. 环境配置

conda install pytorch==2.1.2 torchvision==0.16.2 torchaudio==2.1.2 pytorch-cuda=12.1 -c pytorch -c nvidia

git clone https://github.com/nnanhuang/S3Gaussian.git --recursive
cd S3Gaussian
conda create -n S3Gaussian python=3.9 
conda activate S3Gaussianpip install -r requirements.txt
pip install -e submodules/depth-diff-gaussian-rasterization
pip install -e submodules/simple-knn

2. 数据准备

2.1. 数据下载

# Create the data directory or create a symbolic link to the data directory
mkdir -p [your path]/waymo/raw   
mkdir -p [your path]/waymo/processed 

2.2. 数据处理

python preprocess_main.py --data_root [your path]/waymo/raw --target_dir [your path]/waymo/processed --split training --process_keys images lidar calib pose dynamic_masks --workers 2 --scene_ids 114 700

3. 运行

3.1. 训练

python train.py -s "[your path]/waymo/processed/training/114" --expname "waymo" --model_path "[your path]/waymo/model"

3.2. 推理

参考文献

https://github.com/nnanhuang/S3Gaussian/tree/main?tab=readme-ov-file