Swiftbrush

🚀 SwiftBrush Model Card

SwiftBrush is a one - step text - to - image diffusion model. It offers a novel approach to text - to - image generation through variational score distillation, providing high - quality image outputs.

🚀 Quick Start

This section is not provided in the original README, so it is skipped.

✨ Features

• Text - to - Image Generation: Capable of generating images based on text descriptions.
• One - Step Diffusion: Achieves image generation in a single step with variational score distillation.
• Research - Oriented: Intended for research purposes, and a superior version (SwiftBrush v2) is under development.

📚 Documentation

Note

This model is for research purpose only. We have also been developing a superior version, SwiftBrush v2, and a brief introduction of it is available here.

Contact

Feel free to contact us if you have any questions about the paper!

Thuan Hoang Nguyen v.thuannh5@vinai.io

For more information, please refer to the code repository

License

SwiftBrush is released under [Creative Commons Attribution - NonCommercial - ShareAlike 4.0 International License](https://creativecommons.org/licenses/by - nc - sa/4.0/deed.en).

Citation

If you find SwiftBrush useful or relevant to your research, please kindly cite our papers:

@InProceedings{nguyen2024swiftbrush,
  title={SwiftBrush: One - Step Text - to - Image Diffusion Model with Variational Score Distillation},
  author={Thuan Hoang Nguyen and Anh Tran},
  booktitle = {Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)},
  year={2024}
}

Acknowledgments

We give thanks to Uy Dieu Tran for early discussions as well as providing many helpful comments and suggestions throughout the project. Special thanks to Trung Tuan Dao for valuable feedback and support. Last but not least, we thank Zhengyi Wang, Cheng Lu, Yikai Wang, Fan Bao, Chongxuan Li, Hang Su and Jun Zhu for the work of ProlificDreamer as well as Huggingface team for the diffusers framework.