Multi-Task Image Restoration Algorithm Under Different Weather Influence Factors

Bad weather, such as rain, snow, and fog, will reduce the quality of the image acquired, and it will affect the performance of many related visual fields. The existing researches on image restoration under severe weather either focus on the restoration under a certain kind of weather, which cannot b...

Full description

Saved in:
Bibliographic Details
Main Authors: Lingwen Meng, Mingyong Xin, Junwei Zhang, Fu Zou, Qingqing Zhao
Format: Article
Language:English
Published: IEEE 2024-01-01
Series:IEEE Access
Subjects:
Online Access:https://ieeexplore.ieee.org/document/10810379/
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850051877851365376
author Lingwen Meng
Mingyong Xin
Junwei Zhang
Fu Zou
Qingqing Zhao
author_facet Lingwen Meng
Mingyong Xin
Junwei Zhang
Fu Zou
Qingqing Zhao
author_sort Lingwen Meng
collection DOAJ
description Bad weather, such as rain, snow, and fog, will reduce the quality of the image acquired, and it will affect the performance of many related visual fields. The existing researches on image restoration under severe weather either focus on the restoration under a certain kind of weather, which cannot be generalized under different weather, or need to introduce additional model structures, which increases the burden of practical applications. Therefore, we explores an integrated image restoration framework, which can restore images under different adverse weather effects. Specifically, the model is trained in two stages. In the pre-training stage, a basic general network that can deal with multiple weather is trained by supervised learning. In the fine-tuning stage, soft prompts are introduced to stimulate the model’s ability to cope with different weather, so as to enhance the generalization ability. Specifically, the hidden information of the prompts is explored by low-rank decomposition, and the contras loss is added to prompt the prompts to converge on similar tasks. Furthermore, we address the distribution shift problem by aligning out-of-distribution (OOD) test sample statistics with those of the source data using test-time prompt tuning. Finally, we use the evaluation metrics, PSNR and SSIM, to evaluate the proposed method under four tasks: rain removal, snow removal, fog removal, and raindrop removal. The results demonstrate that the proposed method achieves superior performance compared to the existing state-of-the-art.
format Article
id doaj-art-eba199e3134b4ea785dbf9638ca09c87
institution DOAJ
issn 2169-3536
language English
publishDate 2024-01-01
publisher IEEE
record_format Article
series IEEE Access
spelling doaj-art-eba199e3134b4ea785dbf9638ca09c872025-08-20T02:52:59ZengIEEEIEEE Access2169-35362024-01-011219738919739910.1109/ACCESS.2024.352048410810379Multi-Task Image Restoration Algorithm Under Different Weather Influence FactorsLingwen Meng0https://orcid.org/0009-0003-0880-4927Mingyong Xin1Junwei Zhang2Fu Zou3Qingqing Zhao4Electric Power Research Institute of Guizhou Power Grid Co. Ltd., Guiyang, ChinaElectric Power Research Institute of Guizhou Power Grid Co. Ltd., Guiyang, ChinaElectric Power Research Institute of Guizhou Power Grid Co. Ltd., Guiyang, ChinaElectric Power Research Institute of Guizhou Power Grid Co. Ltd., Guiyang, ChinaElectric Power Research Institute of Guizhou Power Grid Co. Ltd., Guiyang, ChinaBad weather, such as rain, snow, and fog, will reduce the quality of the image acquired, and it will affect the performance of many related visual fields. The existing researches on image restoration under severe weather either focus on the restoration under a certain kind of weather, which cannot be generalized under different weather, or need to introduce additional model structures, which increases the burden of practical applications. Therefore, we explores an integrated image restoration framework, which can restore images under different adverse weather effects. Specifically, the model is trained in two stages. In the pre-training stage, a basic general network that can deal with multiple weather is trained by supervised learning. In the fine-tuning stage, soft prompts are introduced to stimulate the model’s ability to cope with different weather, so as to enhance the generalization ability. Specifically, the hidden information of the prompts is explored by low-rank decomposition, and the contras loss is added to prompt the prompts to converge on similar tasks. Furthermore, we address the distribution shift problem by aligning out-of-distribution (OOD) test sample statistics with those of the source data using test-time prompt tuning. Finally, we use the evaluation metrics, PSNR and SSIM, to evaluate the proposed method under four tasks: rain removal, snow removal, fog removal, and raindrop removal. The results demonstrate that the proposed method achieves superior performance compared to the existing state-of-the-art.https://ieeexplore.ieee.org/document/10810379/Transformer architecturepretrain-finetuningsoft promptscontrast learningdistribution alignment strategy
spellingShingle Lingwen Meng
Mingyong Xin
Junwei Zhang
Fu Zou
Qingqing Zhao
Multi-Task Image Restoration Algorithm Under Different Weather Influence Factors
IEEE Access
Transformer architecture
pretrain-finetuning
soft prompts
contrast learning
distribution alignment strategy
title Multi-Task Image Restoration Algorithm Under Different Weather Influence Factors
title_full Multi-Task Image Restoration Algorithm Under Different Weather Influence Factors
title_fullStr Multi-Task Image Restoration Algorithm Under Different Weather Influence Factors
title_full_unstemmed Multi-Task Image Restoration Algorithm Under Different Weather Influence Factors
title_short Multi-Task Image Restoration Algorithm Under Different Weather Influence Factors
title_sort multi task image restoration algorithm under different weather influence factors
topic Transformer architecture
pretrain-finetuning
soft prompts
contrast learning
distribution alignment strategy
url https://ieeexplore.ieee.org/document/10810379/
work_keys_str_mv AT lingwenmeng multitaskimagerestorationalgorithmunderdifferentweatherinfluencefactors
AT mingyongxin multitaskimagerestorationalgorithmunderdifferentweatherinfluencefactors
AT junweizhang multitaskimagerestorationalgorithmunderdifferentweatherinfluencefactors
AT fuzou multitaskimagerestorationalgorithmunderdifferentweatherinfluencefactors
AT qingqingzhao multitaskimagerestorationalgorithmunderdifferentweatherinfluencefactors