中文
相关论文

相关论文: All but One: Surgical Concept Erasing with Model P…

200 篇论文

Concept erasure in text-to-image diffusion models is crucial for mitigating harmful content, yet existing methods often compromise generative quality. We introduce Semantic Surgery, a novel training-free, zero-shot framework for concept…

计算机视觉与模式识别 · 计算机科学 2025-10-28 Lexiang Xiong , Chengyu Liu , Jingwen Ye , Yan Liu , Yuecong Xu

Concept erasure is extensively utilized in image generation to prevent text-to-image models from generating undesired content. Existing methods can effectively erase narrow concepts that are specific and concrete, such as distinct…

计算机视觉与模式识别 · 计算机科学 2026-03-10 Yuze Cai , Jiahao Lu , Hongxiang Shi , Yichao Zhou , Hong Lu

Text-to-image models suffer from various safety issues that may limit their suitability for deployment. Previous methods have separately addressed individual issues of bias, copyright, and offensive content in text-to-image models. However,…

计算机视觉与模式识别 · 计算机科学 2024-10-25 Rohit Gandikota , Hadas Orgad , Yonatan Belinkov , Joanna Materzyńska , David Bau

Recent works on diffusion models have demonstrated a strong capability for conditioning image generation, e.g., text-guided image synthesis. Such success inspires many efforts trying to use large-scale pre-trained diffusion models for…

计算机视觉与模式识别 · 计算机科学 2025-04-01 Zhixing Zhang , Ligong Han , Arnab Ghosh , Dimitris Metaxas , Jian Ren

Despite their remarkable image generation capabilities, text-to-image diffusion models inadvertently learn inappropriate concepts from vast and unfiltered training data, which leads to various ethical and business risks. Specifically,…

计算机视觉与模式识别 · 计算机科学 2025-02-18 Die Chen , Zhiwen Li , Mingyuan Fan , Cen Chen , Wenmeng Zhou , Yanhao Wang , Yaliang Li

Diffusion models have demonstrated remarkable capability in generating high-quality visual content from textual descriptions. However, since these models are trained on large-scale internet data, they inevitably learn undesirable concepts,…

机器学习 · 计算机科学 2025-02-18 Anh Bui , Khanh Doan , Trung Le , Paul Montague , Tamas Abraham , Dinh Phung

Removing unwanted concepts from large-scale text-to-image (T2I) diffusion models while maintaining their overall generative quality remains an open challenge. This difficulty is especially pronounced in emerging paradigms, such as Stable…

计算机视觉与模式识别 · 计算机科学 2025-01-03 Daiheng Gao , Shilin Lu , Shaw Walters , Wenbo Zhou , Jiaming Chu , Jie Zhang , Bang Zhang , Mengxi Jia , Jian Zhao , Zhaoxin Fan , Weiming Zhang

Concept erasure has emerged as a promising technique for mitigating the risk of harmful content generation in diffusion models by selectively unlearning undesirable concepts. The common principle of previous works to remove a specific…

机器学习 · 计算机科学 2025-05-26 Anh Bui , Trang Vu , Long Vuong , Trung Le , Paul Montague , Tamas Abraham , Junae Kim , Dinh Phung

Large-scale text-to-image (T2I) diffusion models have revolutionized image generation, enabling the synthesis of highly detailed visuals from textual descriptions. However, these models may inadvertently generate inappropriate content, such…

计算机视觉与模式识别 · 计算机科学 2025-03-24 Huiqiang Chen , Tianqing Zhu , Linlin Wang , Xin Yu , Longxiang Gao , Wanlei Zhou

Remarkable progress in text-to-image diffusion models has brought a major concern about potentially generating images on inappropriate or trademarked concepts. Concept erasing has been investigated with the goals of deleting target concepts…

计算机视觉与模式识别 · 计算机科学 2025-07-01 Byung Hyun Lee , Sungjin Lim , Seunggyu Lee , Dong Un Kang , Se Young Chun

Diffusion models have achieved unprecedented success in image generation but pose increasing risks in terms of privacy, fairness, and security. A growing demand exists to \emph{erase} sensitive or harmful concepts (e.g., NSFW content,…

计算机视觉与模式识别 · 计算机科学 2025-10-08 Zixuan Fu , Yan Ren , Finn Carter , Chenyue Wen , Le Ku , Daheng Yu , Emily Davis , Bo Zhang

Concept Erasure, which aims to prevent pretrained text-to-image models from generating content associated with semantic-harmful concepts (i.e., target concepts), is getting increased attention. State-of-the-art methods formulate this task…

计算机视觉与模式识别 · 计算机科学 2025-08-07 Hongxu Chen , Zhen Wang , Taoran Mei , Lin Li , Bowei Zhu , Runshi Li , Long Chen

Erasing concepts from large-scale text-to-image (T2I) diffusion models has become increasingly crucial due to the growing concerns over copyright infringement, offensive content, and privacy violations. In scalable applications,…

计算机视觉与模式识别 · 计算机科学 2026-03-03 Ouxiang Li , Yuan Wang , Xinting Hu , Houcheng Jiang , Yanbin Hao , Fuli Feng

Diffusion models are highly effective at generating high-quality images but pose risks, such as the unintentional generation of NSFW (not safe for work) content. Although various techniques have been proposed to mitigate unwanted influences…

计算机视觉与模式识别 · 计算机科学 2024-11-21 Jing Wu , Trung Le , Munawar Hayat , Mehrtash Harandi

Concept erasure, which fine-tunes diffusion models to remove undesired or harmful visual concepts, has become a mainstream approach to mitigating unsafe or illegal image generation in text-to-image models.However, existing removal methods…

计算机视觉与模式识别 · 计算机科学 2025-12-17 Hao Chen , Yiwei Wang , Songze Li

While generative models produce high-quality images of concepts learned from a large-scale database, a user often wishes to synthesize instantiations of their own concepts (for example, their family, pets, or items). Can we teach a model to…

计算机视觉与模式识别 · 计算机科学 2023-06-21 Nupur Kumari , Bingliang Zhang , Richard Zhang , Eli Shechtman , Jun-Yan Zhu

Recent advances in text-to-image diffusion models enable photorealistic image generation, but they also risk producing malicious content, such as NSFW images. To mitigate risk, concept erasure methods are studied to facilitate the model to…

计算机视觉与模式识别 · 计算机科学 2025-10-07 Ruidong Chen , Honglin Guo , Lanjun Wang , Chenyu Zhang , Weizhi Nie , An-An Liu

Recent advance in text-to-image diffusion models have significantly facilitated the generation of high-quality images, but also raising concerns about the illegal creation of harmful content, such as copyrighted images. Existing concept…

计算机视觉与模式识别 · 计算机科学 2025-01-06 Zihao Wang , Yuxiang Wei , Fan Li , Renjing Pei , Hang Xu , Wangmeng Zuo

Text-to-image diffusion models have shown unprecedented generative capability, but their ability to produce undesirable concepts (e.g.~pornographic content, sensitive identities, copyrighted styles) poses serious concerns for privacy,…

计算机视觉与模式识别 · 计算机科学 2025-05-30 Finn Carter

Text-to-image models encounter safety issues, including concerns related to copyright and Not-Safe-For-Work (NSFW) content. Despite several methods have been proposed for erasing inappropriate concepts from diffusion models, they often…

计算机视觉与模式识别 · 计算机科学 2024-10-29 Chao Gong , Kai Chen , Zhipeng Wei , Jingjing Chen , Yu-Gang Jiang