通过语义关联识别精准擦除文本图像模型中的目标概念
CRCE: Coreference-Retention Concept Erasure in Text-to-Image Diffusion Models
- 利用大模型识别需擦除和需保留的语义相关概念
- 在多种任务中优于现有方法，避免残留或误删
- 适合需要精确内容控制的图像生成场景
文本到图像扩散模型可能生成不期望的内容，需要进行概念擦除。然而，现有方法存在擦除不足（残留目标概念）或过度擦除（误删视觉相似但无关概念）的问题。为此，我们提出CRCE，一种新颖的概念擦除框架，利用大型语言模型识别与目标概念语义相关的应擦除概念，以及应保留的区分性概念。通过显式建模共指与保留概念的语义关系，CRCE实现更精准的概念移除，避免意外擦除。实验表明，CRCE在多种擦除任务上优于现有方法，包括真实物体、人物身份及抽象知识产权特征。构建的数据集CorefConcept和源代码将在论文接受后发布。
原文摘要 · Abstract (English)
Text-to-Image diffusion models can produce undesirable content that necessitates concept erasure. However, existing methods struggle with under-erasure, leaving residual traces of targeted concepts, or over-erasure, mistakenly eliminating unrelated but visually similar concepts. To address these limitations, we introduce CRCE, a novel concept erasure framework that leverages Large Language Models to identify both semantically related concepts that should be erased alongside the target and distinct concepts that should be preserved. By explicitly modelling coreferential and retained concepts semantically, CRCE enables more precise concept removal, without unintended erasure. Experiments demonstrate that CRCE outperforms existing methods on diverse erasure tasks, including real-world object, person identities, and abstract intellectual property characteristics. The constructed dataset CorefConcept and the source code will be release upon acceptance.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。