Continual Learning-Based Unified Model for Unpaired Image Restoration Tasks

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kartheek, Kotha, Chowdary, Lingamaneni Gnanesh, Mukherjee, Snehasis
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915409325719552
author Kartheek, Kotha
Chowdary, Lingamaneni Gnanesh
Mukherjee, Snehasis
author_facet Kartheek, Kotha
Chowdary, Lingamaneni Gnanesh
Mukherjee, Snehasis
contents Restoration of images contaminated by different adverse weather conditions such as fog, snow, and rain is a challenging task due to the varying nature of the weather conditions. Most of the existing methods focus on any one particular weather conditions. However, for applications such as autonomous driving, a unified model is necessary to perform restoration of corrupted images due to different weather conditions. We propose a continual learning approach to propose a unified framework for image restoration. The proposed framework integrates three key innovations: (1) Selective Kernel Fusion layers that dynamically combine global and local features for robust adaptive feature selection; (2) Elastic Weight Consolidation (EWC) to enable continual learning and mitigate catastrophic forgetting across multiple restoration tasks; and (3) a novel Cycle-Contrastive Loss that enhances feature discrimination while preserving semantic consistency during domain translation. Further, we propose an unpaired image restoration approach to reduce the dependance of the proposed approach on the training data. Extensive experiments on standard benchmark datasets for dehazing, desnowing and deraining tasks demonstrate significant improvements in PSNR, SSIM, and perceptual quality over the state-of-the-art.
format Preprint
id arxiv_https___arxiv_org_abs_2507_19184
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Continual Learning-Based Unified Model for Unpaired Image Restoration Tasks
Kartheek, Kotha
Chowdary, Lingamaneni Gnanesh
Mukherjee, Snehasis
Computer Vision and Pattern Recognition
Restoration of images contaminated by different adverse weather conditions such as fog, snow, and rain is a challenging task due to the varying nature of the weather conditions. Most of the existing methods focus on any one particular weather conditions. However, for applications such as autonomous driving, a unified model is necessary to perform restoration of corrupted images due to different weather conditions. We propose a continual learning approach to propose a unified framework for image restoration. The proposed framework integrates three key innovations: (1) Selective Kernel Fusion layers that dynamically combine global and local features for robust adaptive feature selection; (2) Elastic Weight Consolidation (EWC) to enable continual learning and mitigate catastrophic forgetting across multiple restoration tasks; and (3) a novel Cycle-Contrastive Loss that enhances feature discrimination while preserving semantic consistency during domain translation. Further, we propose an unpaired image restoration approach to reduce the dependance of the proposed approach on the training data. Extensive experiments on standard benchmark datasets for dehazing, desnowing and deraining tasks demonstrate significant improvements in PSNR, SSIM, and perceptual quality over the state-of-the-art.
title Continual Learning-Based Unified Model for Unpaired Image Restoration Tasks
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2507.19184