Kiss3DGen: Repurposing Image Diffusion Models for 3D Asset Generation

---

🚩 Features

• [✅] Release inference code.
• [✅] Release model weights.
• [✅] Release huggingface gradio demo. Please try it at HF demo link.
• [✅] Release local gradio demo (more applications). Please try it at Local demo link.

⚙️ Dependencies and Installation

We recommend using Python>=3.10, PyTorch>=2.4.0, and CUDA>=12.1.

conda create --name kiss3dgen python=3.10
conda activate kiss3dgen
pip install -U pip

# Install the correct version of CUDA
conda install cuda -c nvidia/label/cuda-12.1.0

# 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 torchaudio==2.4.0 --index-url https://download.pytorch.org/whl/cu121
pip install xformers==0.0.27.post1

# Install Pytorch3D 
pip install iopath
pip install --no-index --no-cache-dir pytorch3d -f https://dl.fbaipublicfiles.com/pytorch3d/packaging/wheels/py310_cu121_pyt240/download.html

# Install torch-scatter 
pip install torch-scatter -f https://data.pyg.org/whl/torch-2.4.0+cu121.html

# Install other requirements
pip install -r requirements.txt

💫 Inference

Download the pretrained model

Our inference script will download the models automatically. Alternatively, you can run the following command to download them manually, it will put them under the checkpoint/ directory.

# Download the pretrained models
python ./download_models.py

3D Asset Generation

we run it on A800 GPU with 80GB memory, if you only have a smaller GPU, you can change the models' device in the pipeline/pipeline_config/default.yaml file to use two or more smaller memory GPUs.

# Text-to-3D
python ./pipeline/example_text_to_3d.py
# Image-to-3D
python ./pipeline/example_image_to_3d.py
# 3D-to-3D
python ./pipeline/example_3d_to_3d.py

Gradio Demo

Interactive inference: Run your local gradio demo.

python ./app.py

📜 Citation

If you find our work useful for your research or applications, please cite using this BibTeX:

@article{lin2025kiss3dgen,
  title={Kiss3DGen: Repurposing Image Diffusion Models for 3D Asset Generation},
  author={Lin, Jiantao and Yang, Xin and Chen, Meixi and Xu, Yingjie and Yan, Dongyu and Wu, Leyi and Xu, Xinli and Xu, Lie and Zhang, Shunsi and Chen, Ying-Cong},
  journal={arXiv preprint arXiv:2503.01370},
  year={2025}
}

🤗 Acknowledgements

We thank the authors of the following projects for their excellent contributions to 3D generative AI!

• PRM
• FlexGen
• Unique3D
• FluxGym