Six-CD: Benchmarking Concept Removals for Benign Text-to-image Diffusion Models
This work addresses the need for consistent evaluation of safety measures in AI-generated imagery, though it is incremental as it builds on existing concept removal methods.
The authors tackled the problem of evaluating concept removal methods in text-to-image diffusion models to prevent malicious content generation, by introducing a new dataset and metric, resulting in a benchmark that provides insights through experimental observations.
Text-to-image (T2I) diffusion models have shown exceptional capabilities in generating images that closely correspond to textual prompts. However, the advancement of T2I diffusion models presents significant risks, as the models could be exploited for malicious purposes, such as generating images with violence or nudity, or creating unauthorized portraits of public figures in inappropriate contexts. To mitigate these risks, concept removal methods have been proposed. These methods aim to modify diffusion models to prevent the generation of malicious and unwanted concepts. Despite these efforts, existing research faces several challenges: (1) a lack of consistent comparisons on a comprehensive dataset, (2) ineffective prompts in harmful and nudity concepts, (3) overlooked evaluation of the ability to generate the benign part within prompts containing malicious concepts. To address these gaps, we propose to benchmark the concept removal methods by introducing a new dataset, Six-CD, along with a novel evaluation metric. In this benchmark, we conduct a thorough evaluation of concept removals, with the experimental observations and discussions offering valuable insights in the field.