1 paper
Yixiong Yang, Tao Wu, Senmao Li +4
Recent progress in accelerating text-to-image diffusion models enables high-fidelity synthesis within a single denoising step. However, customizing the fast one-step models remains…