CAN: Cascade Augmentations Against Noise for Image Restoration

doi:10.1109/TIP.2025.3595374

	CAN: Cascade Augmentations Against Noise for Image Restoration
	Yan, Yanyang 1,2; Yao, Siyuan 3; Ren, Wenqi 4; Zhang, Rui 5; Guo, Qi 5; Cao, Xiaochun 4
	2025
发表期刊	IEEE TRANSACTIONS ON IMAGE PROCESSING
ISSN	1057-7149
卷号	34 页码:5131-5146
摘要	Image restoration aims to recover the latent clean image from a degraded counterpart. In general, the prevailing state-of-the-art image restoration methods concentrate on solving only a specific degradation type according to the task, e.g., deblurring or deraining. However, if the corresponding well-trained frameworks confront other real-world image corruptions, i.e., the corruptions are not covered in the training phase, and state-of-the-art restoration models will suffer from a lack of generalization ability. We have observed that an image restoration model can be easily confused by noise corruption. Towards improving the robustness of image restoration networks, in this paper, we focus on alleviating the corruption of noise in various image restoration tasks, which is almost inevitable in real-world scenes. To this end, we devise a novel Cascade Augmentation strategy against Noise (CAN) to enhance the robustness of specific image restoration. Specifically, the given degraded images are sequentially augmented from different perspectives, i.e., noise-aware augmentation and model-aware augmentation. The noise-aware augmentation is proposed to enrich the samples by introducing various noise operations. Moreover, to adapt to more unknown corruptions, we propose a novel model-aware augmentation mechanism, which enhances the scalability by exploring useful both spatial and frequency clues with the help of model randomness. It is worth noting that the proposed augmentation scheme is model-agnostic, and it can plug and play into arbitrary state-of-the-art image restoration architectures. In addition, we construct noise corruption benchmark datasets, derived from the validation set of standard image restoration datasets, to assist us in evaluating the robustness of restoration networks. Extensive quantitative and qualitative evaluations demonstrate that the proposed method has strong generalization capability, which can enhance the robustness of various image restoration frameworks when facing diverse noises.
关键词	Image restoration cascade augmentations cascade augmentations noise corruptions noise corruptions
DOI	10.1109/TIP.2025.3595374
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[62302480] ; National Natural Science Foundation of China[62402055] ; National Natural Science Foundation of China[62322216] ; National Natural Science Foundation of China[62172409] ; National Natural Science Foundation of China[62311530686]
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:001554452200008
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/41781
专题	中国科学院计算技术研究所期刊论文_英文
通讯作者	Ren, Wenqi
作者单位	1.Commun Univ China, Sch Data Sci & Media Intelligence, Beijing 100024, Peoples R China 2.Commun Univ China, Key Lab Media Audio & Video, Minist Educ, Beijing 100024, Peoples R China 3.Beijing Univ Posts & Telecommun, Sch Comp Sci, Nat Pilot Software Engn Sch, Beijing 100876, Peoples R China 4.Sun Yat Sen Univ, Sch Cyber Secur, Shenzhen 518107, Peoples R China 5.Chinese Acad Sci, Inst Comp Technol, State Key Lab Proc, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Yan, Yanyang,Yao, Siyuan,Ren, Wenqi,et al. CAN: Cascade Augmentations Against Noise for Image Restoration[J]. IEEE TRANSACTIONS ON IMAGE PROCESSING,2025,34:5131-5146.
APA	Yan, Yanyang,Yao, Siyuan,Ren, Wenqi,Zhang, Rui,Guo, Qi,&Cao, Xiaochun.(2025).CAN: Cascade Augmentations Against Noise for Image Restoration.IEEE TRANSACTIONS ON IMAGE PROCESSING,34,5131-5146.
MLA	Yan, Yanyang,et al."CAN: Cascade Augmentations Against Noise for Image Restoration".IEEE TRANSACTIONS ON IMAGE PROCESSING 34(2025):5131-5146.