Direct Unlearning Optimization for Robust and Safe Text-to-Image Models
This addresses safety risks in text-to-image models for users and developers, offering a robust solution against adversarial attacks, though it is incremental as it builds on existing unlearning techniques.
The paper tackles the problem of text-to-image models generating unsafe content by proposing Direct Unlearning Optimization (DUO), which robustly removes NSFW content while preserving performance on unrelated topics, as demonstrated by defending against state-of-the-art red teaming methods without significant degradation in FID and CLIP scores.
Recent advancements in text-to-image (T2I) models have unlocked a wide range of applications but also present significant risks, particularly in their potential to generate unsafe content. To mitigate this issue, researchers have developed unlearning techniques to remove the model's ability to generate potentially harmful content. However, these methods are easily bypassed by adversarial attacks, making them unreliable for ensuring the safety of generated images. In this paper, we propose Direct Unlearning Optimization (DUO), a novel framework for removing Not Safe For Work (NSFW) content from T2I models while preserving their performance on unrelated topics. DUO employs a preference optimization approach using curated paired image data, ensuring that the model learns to remove unsafe visual concepts while retaining unrelated features. Furthermore, we introduce an output-preserving regularization term to maintain the model's generative capabilities on safe content. Extensive experiments demonstrate that DUO can robustly defend against various state-of-the-art red teaming methods without significant performance degradation on unrelated topics, as measured by FID and CLIP scores. Our work contributes to the development of safer and more reliable T2I models, paving the way for their responsible deployment in both closed-source and open-source scenarios.