Reliable and Efficient Concept Erasure of Text-to-Image Diffusion Models

🌟 🌟 ECCV 2024 | Arxiv | 🤗HuggingFace 🌟 🌟

Authors

Chao Gong*, Kai Chen*, Zhipeng Wei, Jingjing Chen, Yu-Gang Jiang

Fudan University

Code

The code that has been preliminarily organized has been released.

1. Run pip install -r requirements.txt to install the required packages.

2. You can check scripts/ for running scripts.

The edited models of RECE can be found 🤗here.

Erasure Details

For all concepts, the coefficients of Eq.3 are: $\lambda_1=0.1$ and $\lambda_2=0.1$.

The regularization coefficients $\lambda$ are:

1. Nudity and unsafe concepts(I2P concepts), $\lambda=1e-1$.
2. Artistic styles, $\lambda=1e-3$.
3. Difficult objects (e.g., church and garbage truck), $\lambda=1e-3$.
4. Easy objects (e.g., English Springer, golf ball and parachute), $\lambda=1e-1$.
5. For other objects where erasing accuracies reach 0 using UCE, RECE's further erasure is not applied.

We will update the Arxiv version to correct/align the experiment settings.

Citation

If you find our work helpful, please leave us a star and cite our paper.

@article{gong2024reliable,
  title={Reliable and Efficient Concept Erasure of Text-to-Image Diffusion Models},
  author={Gong, Chao and Chen, Kai and Wei, Zhipeng and Chen, Jingjing and Jiang, Yu-Gang},
  journal={arXiv preprint arXiv:2407.12383},
  year={2024}
}

Acknowledgement

Some code is borrowed from UCE.