Equivariant Sampling for Improving Diffusion Model-based Image Restoration

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Wu, Chenxu, Kong, Qingpeng, Zhao, Peiang, Yang, Wendi, Ma, Wenxin, Tang, Fenghe, Jiang, Zihang, Zhou, S. Kevin
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866912706451210240
author Wu, Chenxu
Kong, Qingpeng
Zhao, Peiang
Yang, Wendi
Ma, Wenxin
Tang, Fenghe
Jiang, Zihang
Zhou, S. Kevin
author_facet Wu, Chenxu
Kong, Qingpeng
Zhao, Peiang
Yang, Wendi
Ma, Wenxin
Tang, Fenghe
Jiang, Zihang
Zhou, S. Kevin
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2511_09965
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Equivariant Sampling for Improving Diffusion Model-based Image Restoration
Wu, Chenxu
Kong, Qingpeng
Zhao, Peiang
Yang, Wendi
Ma, Wenxin
Tang, Fenghe
Jiang, Zihang
Zhou, S. Kevin
Computer Vision and Pattern Recognition
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.
title Equivariant Sampling for Improving Diffusion Model-based Image Restoration
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.09965