1 paper
Hyeonyu Kim, Seokhoon Jeong, Seonghee Han +2
Visual editing with diffusion models has made significant progress but often struggles with complex scenarios that textual guidance alone could not adequately describe, highlightin…