AGLLDiff: Guiding Diffusion Models Towards Unsupervised Training-free Real-world Low-light Image Enhancement

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lin, Yunlong, Ye, Tian, Chen, Sixiang, Fu, Zhenqi, Wang, Yingying, Chai, Wenhao, Xing, Zhaohu, Zhu, Lei, Ding, Xinghao
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917730784903168
author Lin, Yunlong
Ye, Tian
Chen, Sixiang
Fu, Zhenqi
Wang, Yingying
Chai, Wenhao
Xing, Zhaohu
Zhu, Lei
Ding, Xinghao
author_facet Lin, Yunlong
Ye, Tian
Chen, Sixiang
Fu, Zhenqi
Wang, Yingying
Chai, Wenhao
Xing, Zhaohu
Zhu, Lei
Ding, Xinghao
contents Existing low-light image enhancement (LIE) methods have achieved noteworthy success in solving synthetic distortions, yet they often fall short in practical applications. The limitations arise from two inherent challenges in real-world LIE: 1) the collection of distorted/clean image pairs is often impractical and sometimes even unavailable, and 2) accurately modeling complex degradations presents a non-trivial problem. To overcome them, we propose the Attribute Guidance Diffusion framework (AGLLDiff), a training-free method for effective real-world LIE. Instead of specifically defining the degradation process, AGLLDiff shifts the paradigm and models the desired attributes, such as image exposure, structure and color of normal-light images. These attributes are readily available and impose no assumptions about the degradation process, which guides the diffusion sampling process to a reliable high-quality solution space. Extensive experiments demonstrate that our approach outperforms the current leading unsupervised LIE methods across benchmarks in terms of distortion-based and perceptual-based metrics, and it performs well even in sophisticated wild degradation.
format Preprint
id arxiv_https___arxiv_org_abs_2407_14900
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle AGLLDiff: Guiding Diffusion Models Towards Unsupervised Training-free Real-world Low-light Image Enhancement
Lin, Yunlong
Ye, Tian
Chen, Sixiang
Fu, Zhenqi
Wang, Yingying
Chai, Wenhao
Xing, Zhaohu
Zhu, Lei
Ding, Xinghao
Computer Vision and Pattern Recognition
Existing low-light image enhancement (LIE) methods have achieved noteworthy success in solving synthetic distortions, yet they often fall short in practical applications. The limitations arise from two inherent challenges in real-world LIE: 1) the collection of distorted/clean image pairs is often impractical and sometimes even unavailable, and 2) accurately modeling complex degradations presents a non-trivial problem. To overcome them, we propose the Attribute Guidance Diffusion framework (AGLLDiff), a training-free method for effective real-world LIE. Instead of specifically defining the degradation process, AGLLDiff shifts the paradigm and models the desired attributes, such as image exposure, structure and color of normal-light images. These attributes are readily available and impose no assumptions about the degradation process, which guides the diffusion sampling process to a reliable high-quality solution space. Extensive experiments demonstrate that our approach outperforms the current leading unsupervised LIE methods across benchmarks in terms of distortion-based and perceptual-based metrics, and it performs well even in sophisticated wild degradation.
title AGLLDiff: Guiding Diffusion Models Towards Unsupervised Training-free Real-world Low-light Image Enhancement
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2407.14900