Structural residual learning for single image rain removal

To alleviate the adverse effect of rain streaks in image processing tasks, CNN-based single image rain removal methods have been recently proposed. However, the performance of these deep learning methods largely relies on the covering range of rain shapes contained in the pre-collected training rain...

Celý popis

Uložené v:
Podrobná bibliografia
Vydané v:Knowledge-based systems Ročník 213; s. 106595
Hlavní autori: Wang, Hong, Wu, Yichen, Xie, Qi, Zhao, Qian, Liang, Yong, Zhang, Shijun, Meng, Deyu
Médium: Journal Article
Jazyk:English
Vydavateľské údaje: Amsterdam Elsevier B.V 15.02.2021
Elsevier Science Ltd
Predmet:
Computer architecture
Deep learning
Experiments
Generalization
Generalization performance
Image processing
Interpretability
Learning
Multi-scale convolutional sparse coding
Networks
Prior knowledge
Rain
Regularization
Side effects
Single image deraining
Training
Deep learning
Generalization performance
Single image deraining
Interpretability
Multi-scale convolutional sparse coding
ISSN:0950-7051, 1872-7409
On-line prístup:Získať plný text
Tagy: Pridať tag
Žiadne tagy, Buďte prvý, kto otaguje tento záznam!
Popis
Shrnutí:To alleviate the adverse effect of rain streaks in image processing tasks, CNN-based single image rain removal methods have been recently proposed. However, the performance of these deep learning methods largely relies on the covering range of rain shapes contained in the pre-collected training rainy-clean image pairs. This makes them easily trapped into the overfitting-to-the-training-samples issue and cannot finely generalize to practical rainy images with complex and diverse rain streaks. Against this generalization issue, this study proposes a new network architecture by enforcing the output residual of the network possess intrinsic rain structures. Such a structural residual setting guarantees the rain layer extracted by the network finely comply with the prior knowledge of general rain streaks, and thus regulates sound rain shapes capable of being well extracted from rainy images in both training and predicting stages. Such a general regularization function naturally leads to both its better training accuracy and testing generalization capability even for those non-seen rain configurations. Such superiority is comprehensively substantiated by experiments implemented on synthetic and real datasets both visually and quantitatively as compared with current state-of-the-art methods.
Bibliografia:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0950-7051
1872-7409
DOI:10.1016/j.knosys.2020.106595