1 paper
Junyeong Ahn, Seojin Yoon, Sungyong Baik
As text-to-image diffusion models grow increasingly prevalent, the ability to remove specific concepts-mostly explicit content and many copyrighted characters or styles-has become…