Protective Perturbations against Unauthorized Data Usage in Diffusion-based Image Generation
It addresses privacy and intellectual property issues for users of diffusion models, but is incremental as it systematizes existing knowledge.
The paper surveys protective perturbation methods that prevent unauthorized data usage in diffusion-based image generation by applying adversarial attacks to customization samples, and proposes an evaluation framework to advance research in this area.
Diffusion-based text-to-image models have shown immense potential for various image-related tasks. However, despite their prominence and popularity, customizing these models using unauthorized data also brings serious privacy and intellectual property issues. Existing methods introduce protective perturbations based on adversarial attacks, which are applied to the customization samples. In this systematization of knowledge, we present a comprehensive survey of protective perturbation methods designed to prevent unauthorized data usage in diffusion-based image generation. We establish the threat model and categorize the downstream tasks relevant to these methods, providing a detailed analysis of their designs. We also propose a completed evaluation framework for these perturbation techniques, aiming to advance research in this field.