A Lightweight Dense Connected Approach with Attention on Single Image Super-Resolution
In recent years, neural networks for single image super-resolution (SISR) have applied more profound and deeper network structures to extract extra image details, which brings difficulties in model training. To deal with deep model training problems, researchers utilize dense skip connections to pro...
Main Authors: | , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2021-05-01
|
Series: | Electronics |
Subjects: | |
Online Access: | https://www.mdpi.com/2079-9292/10/11/1234 |
id |
doaj-2f47a8437eab41c6a202c654fc4d1be5 |
---|---|
record_format |
Article |
spelling |
doaj-2f47a8437eab41c6a202c654fc4d1be52021-06-01T00:47:17ZengMDPI AGElectronics2079-92922021-05-01101234123410.3390/electronics10111234A Lightweight Dense Connected Approach with Attention on Single Image Super-ResolutionLei Zha0Yu Yang1Zicheng Lai2Ziwei Zhang3Juan Wen4College of Information and Electrical Engineering, China Agricultural University, Beijing 100083, ChinaCollege of Information and Electrical Engineering, China Agricultural University, Beijing 100083, ChinaSchool of Advanced Materials and Nanotechnology, Xidian University, Xian 710061, ChinaCollege of Information and Electrical Engineering, China Agricultural University, Beijing 100083, ChinaCollege of Information and Electrical Engineering, China Agricultural University, Beijing 100083, ChinaIn recent years, neural networks for single image super-resolution (SISR) have applied more profound and deeper network structures to extract extra image details, which brings difficulties in model training. To deal with deep model training problems, researchers utilize dense skip connections to promote the model’s feature representation ability by reusing deep features of different receptive fields. Benefiting from the dense connection block, SRDensenet has achieved excellent performance in SISR. Despite the fact that the dense connected structure can provide rich information, it will also introduce redundant and useless information. To tackle this problem, in this paper, we propose a Lightweight Dense Connected Approach with Attention for Single Image Super-Resolution (LDCASR), which employs the attention mechanism to extract useful information in channel dimension. Particularly, we propose the recursive dense group (RDG), consisting of Dense Attention Blocks (DABs), which can obtain more significant representations by extracting deep features with the aid of both dense connections and the attention module, making our whole network attach importance to learning more advanced feature information. Additionally, we introduce the group convolution in DABs, which can reduce the number of parameters to 0.6 M. Extensive experiments on benchmark datasets demonstrate the superiority of our proposed method over five chosen SISR methods.https://www.mdpi.com/2079-9292/10/11/1234dense skip connectionssingle image super-resolutiondeep featureschannel attention mechanism |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Lei Zha Yu Yang Zicheng Lai Ziwei Zhang Juan Wen |
spellingShingle |
Lei Zha Yu Yang Zicheng Lai Ziwei Zhang Juan Wen A Lightweight Dense Connected Approach with Attention on Single Image Super-Resolution Electronics dense skip connections single image super-resolution deep features channel attention mechanism |
author_facet |
Lei Zha Yu Yang Zicheng Lai Ziwei Zhang Juan Wen |
author_sort |
Lei Zha |
title |
A Lightweight Dense Connected Approach with Attention on Single Image Super-Resolution |
title_short |
A Lightweight Dense Connected Approach with Attention on Single Image Super-Resolution |
title_full |
A Lightweight Dense Connected Approach with Attention on Single Image Super-Resolution |
title_fullStr |
A Lightweight Dense Connected Approach with Attention on Single Image Super-Resolution |
title_full_unstemmed |
A Lightweight Dense Connected Approach with Attention on Single Image Super-Resolution |
title_sort |
lightweight dense connected approach with attention on single image super-resolution |
publisher |
MDPI AG |
series |
Electronics |
issn |
2079-9292 |
publishDate |
2021-05-01 |
description |
In recent years, neural networks for single image super-resolution (SISR) have applied more profound and deeper network structures to extract extra image details, which brings difficulties in model training. To deal with deep model training problems, researchers utilize dense skip connections to promote the model’s feature representation ability by reusing deep features of different receptive fields. Benefiting from the dense connection block, SRDensenet has achieved excellent performance in SISR. Despite the fact that the dense connected structure can provide rich information, it will also introduce redundant and useless information. To tackle this problem, in this paper, we propose a Lightweight Dense Connected Approach with Attention for Single Image Super-Resolution (LDCASR), which employs the attention mechanism to extract useful information in channel dimension. Particularly, we propose the recursive dense group (RDG), consisting of Dense Attention Blocks (DABs), which can obtain more significant representations by extracting deep features with the aid of both dense connections and the attention module, making our whole network attach importance to learning more advanced feature information. Additionally, we introduce the group convolution in DABs, which can reduce the number of parameters to 0.6 M. Extensive experiments on benchmark datasets demonstrate the superiority of our proposed method over five chosen SISR methods. |
topic |
dense skip connections single image super-resolution deep features channel attention mechanism |
url |
https://www.mdpi.com/2079-9292/10/11/1234 |
work_keys_str_mv |
AT leizha alightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution AT yuyang alightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution AT zichenglai alightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution AT ziweizhang alightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution AT juanwen alightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution AT leizha lightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution AT yuyang lightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution AT zichenglai lightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution AT ziweizhang lightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution AT juanwen lightweightdenseconnectedapproachwithattentiononsingleimagesuperresolution |
_version_ |
1721413843913015296 |