1 paper · 1 filter
Xinyu Liu, Yingqing He, Lanqing Guo +10
The potential for higher-resolution image generation using pretrained diffusion models is immense, yet these models often struggle with issues of object repetition and structural a…