Equivariant Sampling for Improving Diffusion Model-based Image Restoration

TL;DR

This paper introduces EquS, a novel diffusion model-based image restoration method that uses equivariant sampling and a timestep-aware schedule to improve performance and efficiency without extra costs.

Contribution

The paper proposes EquS, a new approach that leverages equivariant information and a timestep-aware schedule to enhance diffusion model-based image restoration.

Findings

Significant performance improvements on benchmark datasets.

Compatibility with existing problem-agnostic DMIR methods.

No additional computational costs incurred.

Abstract

Recent advances in generative models, especially diffusion models, have significantly improved image restoration (IR) performance. However, existing problem-agnostic diffusion model-based image restoration (DMIR) methods face challenges in fully leveraging diffusion priors, resulting in suboptimal performance. In this paper, we address the limitations of current problem-agnostic DMIR methods by analyzing their sampling process and providing effective solutions. We introduce EquS, a DMIR method that imposes equivariant information through dual sampling trajectories. To further boost EquS, we propose the Timestep-Aware Schedule (TAS) and introduce EquS$^+$. TAS prioritizes deterministic steps to enhance certainty and sampling efficiency. Extensive experiments on benchmarks demonstrate that our method is compatible with previous problem-agnostic DMIR methods and significantly boosts their performance without increasing computational costs. Our code is available at https://github.com/FouierL/EquS.