Denoising Diffusion models have exhibited remarkable capabilities in image generation. However, generating high-quality samples requires a large number of iterations. Knowledge distillation for diffusion models is an effective method to address this limitation with a shortened sampling process but causes degraded generative quality. Based on our analysis with bias-variance decomposition and experimental observations, we attribute the degradation to the spatial fitting error occurring in the training of both the teacher and student model. Accordingly, we propose Spatial Fitting-Error Reduction Distillation model (SFERD). SFERD utilizes attention guidance from the teacher model and a designed semantic gradient predictor to reduce the student's fitting error. Empirically, our proposed model facilitates high-quality sample generation in a few function evaluations. We achieve an FID of 5.31 on CIFAR-10 and 9.39 on ImageNet 64×64 with only one step, outperforming existing diffusion methods. Our study provides a new perspective on diffusion distillation by highlighting the intrinsic denoising ability of models. Project link: \url{https://github.com/Sainzerjj/SFERD}.
@article{arxiv.2311.03830,
title = {Reducing Spatial Fitting Error in Distillation of Denoising Diffusion Models},
author = {Shengzhe Zhou and Zejian Lee and Shengyuan Zhang and Lefan Hou and Changyuan Yang and Guang Yang and Zhiyuan Yang and Lingyun Sun},
journal= {arXiv preprint arXiv:2311.03830},
year = {2023}
}