L1 model-driven recursive multi-scale denoising network for image super-resolution

Most existing deep learning based single-image super-resolution (SISR) methods mainly improve the reconstruction performances from the perspective of data-driven, i.e., widening or deepening the networks according to the huge scale of the training data. However, it will bring a huge amount of weight...

Full description

Saved in:
Bibliographic Details
Published in:Knowledge-based systems Vol. 225; p. 1
Main Authors: Sun, Zhongfan, Zhao, Jianwei, Zhou, Zhenghua, Gao, Qingqing
Format: Journal Article
Language:English
Published: Amsterdam Elsevier B.V 05.08.2021
Elsevier Science Ltd
Subjects:
[formula omitted] model-driven
Algorithms
Data
Deep learning
Denoising
Denoising network
Fidelity
Grammatical aspect
Image resolution
Iteration algorithm
Iterative algorithms
Iterative methods
Machine learning
Majorization–minimization algorithm
Minimization
Networks
Noise reduction
Optimization
Outliers (statistics)
Prior knowledge
Recursion
Soft thresholding operator
Super-resolution
Deep learning
Denoising network
Super-resolution
Iteration algorithm
ℓ1 model-driven
Majorization–minimization algorithm
Soft thresholding operator
ISSN:0950-7051, 1872-7409
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Most existing deep learning based single-image super-resolution (SISR) methods mainly improve the reconstruction performances from the perspective of data-driven, i.e., widening or deepening the networks according to the huge scale of the training data. However, it will bring a huge amount of weights and biases, and cost the expensive computations. Recently, some people have proposed a new frame for designing the deep networks according to the algorithms deduced from the ℓ2-optimization problem. But they did not consider the case with outliers. Since ℓ1-norm can describe the sparsity of the outliers better than ℓ2-norm, we propose an effective deep network designed according to the new algorithm deduced from the ℓ1-optimization problem. In our proposed method, an effective iterative algorithm for the ℓ1 reconstructed optimization problem is deduced based on the split Bregman algorithm, majorization–minimization algorithm, and soft thresholding operator. Then according to the deduced iterative algorithm, an effective deep network, named ℓ1 Model-Driven Recursive Multi-Scale Denoising Network (ℓ1-MRMDN), is designed. Due to the iteration form of the deduced algorithm, the proposed ℓ1-MRMDN contains an inner recursion and an outer recursion. Therefore, our proposed method can not only relieve its sensitiveness to the outliers because of the ℓ1 data fidelity term, but also avoid designing the deep network blindly via the guidance of prior knowledge. Extensive experimental results illustrate that our proposed method is superior to some related popular SISR methods.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ISSN:0950-7051
1872-7409
DOI:10.1016/j.knosys.2021.107115