SafeCtrl: Region-Aware Safety Control for Text-to-Image Diffusion via Detect-Then-Suppress
This addresses safety challenges in text-to-image generation for users and developers, offering a more robust solution compared to existing methods, though it is incremental in improving upon prior safety interventions.
The paper tackles the problem of generating harmful content in text-to-image diffusion models by proposing SafeCtrl, a region-aware safety control framework that localizes and suppresses unsafe regions, achieving a superior trade-off between safety and fidelity with improved resilience against adversarial attacks.
The widespread deployment of text-to-image diffusion models is significantly challenged by the generation of visually harmful content, such as sexually explicit content, violence, and horror imagery. Common safety interventions, ranging from input filtering to model concept erasure, often suffer from two critical limitations: (1) a severe trade-off between safety and context preservation, where removing unsafe concepts degrades the fidelity of the safe content, and (2) vulnerability to adversarial attacks, where safety mechanisms are easily bypassed. To address these challenges, we propose SafeCtrl, a Region-Aware safety control framework operating on a Detect-Then-Suppress paradigm. Unlike global safety interventions, SafeCtrl first employs an attention-guided Detect module to precisely localize specific risk regions. Subsequently, a localized Suppress module, optimized via image-level Direct Preference Optimization (DPO), neutralizes harmful semantics only within the detected areas, effectively transforming unsafe objects into safe alternatives while leaving the surrounding context intact. Extensive experiments across multiple risk categories demonstrate that SafeCtrl achieves a superior trade-off between safety and fidelity compared to state-of-the-art methods. Crucially, our approach exhibits improved resilience against adversarial prompt attacks, offering a precise and robust solution for responsible generation.