1 paper · 1 filter
Lingyun Zhang, Yu Xie, Zhongli Fang +2
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,…