Comprehensive Evaluation and Analysis for NSFW Concept Erasure in Text-to-Image Diffusion Models

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Chen, Die, Li, Zhiwen, Chen, Cen, Xie, Yuexiang, Li, Xiaodan, Ye, Jinyan, Chen, Yingda, Li, Yaliang
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866915570877726720
author Chen, Die
Li, Zhiwen
Chen, Cen
Xie, Yuexiang
Li, Xiaodan
Ye, Jinyan
Chen, Yingda
Li, Yaliang
author_facet Chen, Die
Li, Zhiwen
Chen, Cen
Xie, Yuexiang
Li, Xiaodan
Ye, Jinyan
Chen, Yingda
Li, Yaliang
contents Text-to-image diffusion models have gained widespread application across various domains, demonstrating remarkable creative potential. However, the strong generalization capabilities of diffusion models can inadvertently lead to the generation of not-safe-for-work (NSFW) content, posing significant risks to their safe deployment. While several concept erasure methods have been proposed to mitigate the issue associated with NSFW content, a comprehensive evaluation of their effectiveness across various scenarios remains absent. To bridge this gap, we introduce a full-pipeline toolkit specifically designed for concept erasure and conduct the first systematic study of NSFW concept erasure methods. By examining the interplay between the underlying mechanisms and empirical observations, we provide in-depth insights and practical guidance for the effective application of concept erasure methods in various real-world scenarios, with the aim of advancing the understanding of content safety in diffusion models and establishing a solid foundation for future research and development in this critical area.
format Preprint
id arxiv_https___arxiv_org_abs_2505_15450
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Comprehensive Evaluation and Analysis for NSFW Concept Erasure in Text-to-Image Diffusion Models
Chen, Die
Li, Zhiwen
Chen, Cen
Xie, Yuexiang
Li, Xiaodan
Ye, Jinyan
Chen, Yingda
Li, Yaliang
Computer Vision and Pattern Recognition
Text-to-image diffusion models have gained widespread application across various domains, demonstrating remarkable creative potential. However, the strong generalization capabilities of diffusion models can inadvertently lead to the generation of not-safe-for-work (NSFW) content, posing significant risks to their safe deployment. While several concept erasure methods have been proposed to mitigate the issue associated with NSFW content, a comprehensive evaluation of their effectiveness across various scenarios remains absent. To bridge this gap, we introduce a full-pipeline toolkit specifically designed for concept erasure and conduct the first systematic study of NSFW concept erasure methods. By examining the interplay between the underlying mechanisms and empirical observations, we provide in-depth insights and practical guidance for the effective application of concept erasure methods in various real-world scenarios, with the aim of advancing the understanding of content safety in diffusion models and establishing a solid foundation for future research and development in this critical area.
title Comprehensive Evaluation and Analysis for NSFW Concept Erasure in Text-to-Image Diffusion Models
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2505.15450