注意，requirements.txt 中少了 onnxruntime，已补上
We recommend using Python>=3.10, PyTorch>=2.1.0, and CUDA>=12.1.
conda create --name freesplatter python=3.10
conda activate freesplatter
pip install -U pip
# Install PyTorch and xformers
# You may need to install another xformers version if you use a different PyTorch version
pip install torch==2.4.0 torchvision==0.19.0 --index-url https://download.pytorch.org/whl/cu121
pip install xformers==0.0.27.post2
# Install other requirements
pip install -r requirements.txt网络问题，requirements.txt 中的一些包要手动 clone 然后安装。
cd submodule
rm -r diff-gaussian-rasterization
rm -r diff-surfel-rasterization
rm -r nvdiffrast
git clone --recurse-submodules [email protected]:ashawkey/diff-gaussian-rasterization.git
git clone --recurse-submodules [email protected]:hbb1/diff-surfel-rasterization.git
git clone --recurse-submodules [email protected]:NVlabs/nvdiffrast.gitWe provide the following pretrained models:
| Model | Description | #Params | Download |
|---|---|---|---|
| FreeSplatter-O | Object-level reconstruction model | 306M | Download |
| FreeSplatter-O-2dgs | Object-level reconstruction model using 2DGS (finetuned from FreeSplatter-O) | 306M | Download |
| FreeSplatter-S | Scene-level reconstruction model | 306M | Download |
We recommand to start a gradio demo in your local machine, simply run:
python app.py