Groot: Adversarial Testing for Generative Text-to-Image Models with Tree-based Semantic Transformation

arXiv:2402.12100v1 Announce Type: new
Abstract: With the prevalence of text-to-image generative models, their safety becomes a critical concern. adversarial testing techniques have been developed to probe whether such models can be prompted to produce Not-Safe-For-Work (NSFW) content. However, existing solutions face several challenges, including low success rate and inefficiency. We introduce Groot, the first automated framework leveraging tree-based semantic transformation for adversarial testing of text-to-image models. Groot employs semantic decomposition and sensitive element drowning strategies in conjunction with LLMs to …

abstract adversarial arxiv challenges cs.ai cs.cl cs.cr cs.se face generative generative models groot image low nsfw probe rate safety semantic solutions success testing text text-to-image transformation tree type work