Universal Prompt Optimizer for Safe Text-to-Image Generation

arXiv:2402.10882v1 Announce Type: new
Abstract: Text-to-Image (T2I) models have shown great performance in generating images based on textual prompts. However, these models are vulnerable to unsafe input to generate unsafe content like sexual, harassment and illegal-activity images. Existing studies based on image checker, model fine-tuning and embedding blocking are impractical in real-world applications. Hence, \textit{we propose the first universal prompt optimizer for safe T2I generation in black-box scenario}. We first construct a dataset consisting of toxic-clean prompt pairs by GPT-3.5 …

abstract applications arxiv blocking cs.cl cs.cv embedding fine-tuning generate harassment image image generation images model fine-tuning performance prompt prompts studies text text-to-image textual type vulnerable world