Deep Equilibrium Diffusion Restoration with Parallel Sampling

arXiv:2311.11600v2 Announce Type: replace
Abstract: Diffusion model-based image restoration (IR) aims to use diffusion models to recover high-quality (HQ) images from degraded images, achieving promising performance. Due to the inherent property of diffusion models, most existing methods need long serial sampling chains to restore HQ images step-by-step, resulting in expensive sampling time and high computation costs. Moreover, such long sampling chains hinder understanding the relationship between inputs and restoration results since it is hard to compute the gradients in the …

abstract arxiv cs.cv diffusion diffusion model diffusion models equilibrium image image restoration images performance property quality restore sampling step-by-step type