Selective Guidance: Are All the Denoising Steps of Guided Diffusion Important?

05/16/2023

∙

by   Pareesa Ameneh Golnari, et al.

∙

0

∙

share

This study examines the impact of optimizing the Stable Diffusion (SD) guided inference pipeline. We propose optimizing certain denoising steps by limiting the noise computation to conditional noise and eliminating unconditional noise computation, thereby reducing the complexity of the target iterations by 50 Additionally, we demonstrate that later iterations of the SD are less sensitive to optimization, making them ideal candidates for applying the suggested optimization. Our experiments show that optimizing the last 20 denoising loop iterations results in an 8.2 almost no perceivable changes to the human eye. Furthermore, we found that by extending the optimization to 50 inference time by approximately 20.3 images.