Formnet: Formatted learning for image restoration
Publication Type
Journal Article
Publication Date
1-2020
Abstract
In this paper, we propose a deep CNN to tackle the image restoration problem by learning formatted information. Previous deep learning based methods directly learn the mapping from corrupted images to clean images, and may suffer from the gradient exploding/vanishing problems of deep neural networks. We propose to address the image restoration problem by learning the structured details and recovering the latent clean image together, from the shared information between the corrupted image and the latent image. In addition, instead of learning the pure difference (corruption), we propose to add a residual formatting layer and an adversarial block to format the information to structured one, which allows the network to converge faster and boosts the performance. Furthermore, we propose a cross-level loss net to ensure both pixel-level accuracy and semantic-level visual quality. Evaluations on public datasets show that the proposed method performs favorably against existing approaches quantitatively and qualitatively.
Keywords
Image restoration, format, residual, GAN, CNN
Discipline
English Language and Literature
Research Areas
Information Systems and Management
Publication
IEEE Transactions on Image Processing
Volume
29
First Page
6302
Last Page
6314
ISSN
1057-7149
Identifier
10.1109/TIP.2020.2990603
Publisher
Institute of Electrical and Electronics Engineers
Citation
JIAO, Jianbo; TU, Wei-Chih; LIU, Ding; HE, Shengfeng; LAU, Rynson W. H.; and HUANG, Thomas S..
Formnet: Formatted learning for image restoration. (2020). IEEE Transactions on Image Processing. 29, 6302-6314.
Available at: https://ink.library.smu.edu.sg/sis_research/7860
Additional URL
https://doi.org/10.1109/TIP.2020.2990603