Channel Consistency Prior and Self-Reconstruction Strategy Based Unsupervised Image Deraining

2025-03-24

AI Summary

To address the scarcity of paired training data in real-world scenarios and the poor generalization of existing supervised deraining methods, this paper proposes CSUD—the first unsupervised image deraining framework for unpaired data. Methodologically, CSUD introduces two key innovations: (1) a Channel Consistency Prior (CCP), which enforces cross-channel structural constraints to improve rain streak modeling fidelity; and (2) a Self-Reconstruction (SR) strategy that suppresses redundant information transfer in the generator, thereby enhancing model generalization. Crucially, CSUD requires no ground-truth clean–rainy image pairs; instead, it synthesizes high-quality pseudo-paired samples from unlabeled data and jointly optimizes the model via generative adversarial training and a novel Channel Consistency Loss (CCLoss). Extensive experiments demonstrate that CSUD achieves state-of-the-art performance on multiple synthetic and real-world rainy image datasets, significantly improving both deraining quality and cross-domain generalization capability.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Computer Vision: Diffusion Models for Vision
• Cognitive Modeling & Cognitive Systems: Computational Creativity

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Web Mining and Content Analysis: Large pretrained models with web data

Abstract

Recently, deep image deraining models based on paired datasets have made a series of remarkable progress. However, they cannot be well applied in real-world applications due to the difficulty of obtaining real paired datasets and the poor generalization performance. In this paper, we propose a novel Channel Consistency Prior and Self-Reconstruction Strategy Based Unsupervised Image Deraining framework, CSUD, to tackle the aforementioned challenges. During training with unpaired data, CSUD is capable of generating high-quality pseudo clean and rainy image pairs which are used to enhance the performance of deraining network. Specifically, to preserve more image background details while transferring rain streaks from rainy images to the unpaired clean images, we propose a novel Channel Consistency Loss (CCLoss) by introducing the Channel Consistency Prior (CCP) of rain streaks into training process, thereby ensuring that the generated pseudo rainy images closely resemble the real ones. Furthermore, we propose a novel Self-Reconstruction (SR) strategy to alleviate the redundant information transfer problem of the generator, further improving the deraining performance and the generalization capability of our method. Extensive experiments on multiple synthetic and real-world datasets demonstrate that the deraining performance of CSUD surpasses other state-of-the-art unsupervised methods and CSUD exhibits superior generalization capability.

Problem

• Unsupervised image deraining without paired datasets
• Preserving background details while transferring rain streaks
• Improving generalization in real-world deraining applications

Innovation

• Unsupervised framework using Channel Consistency Prior
• Self-Reconstruction Strategy reduces redundant information
• Generates pseudo image pairs for enhanced deraining