计算机科学

计算机科学 ›› 2023, Vol. 50 ›› Issue (6A): 220600240-5.doi: 10.11896/jsjkx.220600240

• 图像处理&多媒体技术 • 上一篇    下一篇

结构化混合注意力网络的图像超分辨率重建

喻九阳, 张德安, 戴耀南, 胡天豪, 夏文凤   

  1. 湖北省绿色化工装备工程技术研究中心 武汉 430205;
    武汉工程大学机电工程学院 武汉 430205
  • 出版日期:2023-06-10 发布日期:2023-06-12
  • 通讯作者: 戴耀南(dyn1121758919@163.com)
  • 作者简介:(yjy@wit.edu.cn)
  • 基金资助:
    湖北省重点研发计划(2020BAB030)

Image Super-resolution Reconstruction Based on Structured Fusion Attention Network

YU Jiuyang, ZHANG Dean, DAI Yaonan, HU Tianhao, XIA Wenfeng   

  1. Hubei Green Chemical Equipment Engineering Technology Research Center,Wuhan 430205,China;
    School of Mechanical and Electrical Engineering,Wuhan Institute of Technology,Wuhan 430205,China
  • Online:2023-06-10 Published:2023-06-12
  • About author:YU Jiuyang,born in 1963,master,se-cond-level professor.His main research interests include chemical machinery process equipment control,oil and gas chemical pipeline robots. DAI Yaonan,born in 1993,Ph.D,lectu-rer.His main research interests include chemical pipeline robotmotion control and mobile robot image recognition.
  • Supported by:
    Key R&D Program of Hubei Province(2020BAB030).

PDF (PC)

285

摘要： 针对现有图像超分辨模型存在特征提取能力弱、模型参数量较复杂等问题,提出了一种结构化混合注意力网络的图像超分辨率重建模型,该模型在提高图像超分辨率重建效果的同时降低了模型的参数量。首先,对编码器进行结构化处理,通过通道数量的不同来提取更多的图像特征。其次,对编码器的输出特性进行注意力网络混合重组,从而加强图像的特征特性。最后,采用残差方式将输入的浅层图像特征直接与强化特征相混合,降低网络的参数量。实验结果表明,在公共数据集及不同放大倍率的前提下,文中构建模型的PSNR值和SSIM值基本是最优的,且网络结构的参数量较低,较好地平衡了图像超分辨率重建过程中性能和参数复杂度间的关系。

关键词: 图像处理, 超分辨率, 结构化残差, 混合注意力, 低模型参数

Abstract: Aiming at the problems of weak feature extraction ability and complex model parameters in existing image super-resolution models,an image super-resolution reconstruction model based on structured hybrid attention network is proposed.This model can reduce the number of parameter while improving the super-resolution reconstruction effect.First,the encoder is structured to extract more image features through the difference in the number of channels.Second,the attention network hybrid reorganization is performed on the output features of the encoder to enhance the feature characteristics of the image.Finally,a residual method is used to directly mix the input shallow image features with the enhanced features to reduce the amount of network parameters.Experimental results show that under the premise of public data sets and different magnifications,the PSNR value and SSIM value of the proposed model are basically optimal,and the parameter amount of the network structure is low,which better balances the relationship between performance and parameter complexity in the process of image super-resolution reconstruction.

Key words: Image processing, Super-resolution, Structured residuals, Fused attention, Low model parameters

中图分类号: 

  • TP391
[1]CHEN Q,SONG H,YU J,et al.Current development and appli-cations of super-resolution ultrasound imaging[J].Sensors,2021,21(7):2417.
[2]LAN R,SUN L,LIU Z,et al.Cascading and enhanced residualnetworks for accurate single-image super-resolution[J].IEEE Transactions on Cybernetics,2020,51(1):115-125.
[3]HU X Y,GUO H R,ZHU R.Image Super-Resolution Recon-struction Based on Hybrid Deep Convolutional Networks[J].Computer Applications,2020,40(7):2069-2076.
[4]YANG X,GUO Y,LI Z,et al.Image super-resolution network based on a multi-branch attention mechanism[J].Signal,Image and Video Processing,2021,15(7):1397-1405.
[5]WANG X,GU Y,GAO X,et al.Dual residual attention module network for single image super resolution[J].Neurocomputing,2019,364:269-279.
[6]KIM J,LEE J K,LEE K M.Accurate image super-resolutionusing very deep convolutional networks[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2016:1646-1654.
[7]LIU J,ZHANG W,TANG Y,et al.Residual feature aggregation network for image super-resolution[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.2020:2359-2368.
[8]ZHANG Y,TIAN Y,KONGY,et al.Residual dense network for image restoration[J].IEEE Transactions on Pattern Analysis and Machine Intelligence,2020,43(7):2480-2495.
[9]DAI Y,YU J,HU T,et al.Structured Fusion Attention Network for Image Super-Resolution Reconstruction[J].IEEE Access,2022,10:31896-31906.
[10]HE K,ZHANG X,REN S,et al.Delving deep into rectifiers:Surpassing human-level performance on imagenet classification[C]//Proceedings of the IEEE international conference on computer vision.2015:1026-1034.
[11]CHOLLET F.Xception:Deep learning with depthwise separable convolutions[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2017:1251-1258.
[12]LI J,FANG F,MEI K,et al.Multi-scale residual network for image super-resolution[C]//Proceedings of the European Conference on Computer Vision(ECCV).2018:517-532.
[13]CHEN H,GU J,ZHANG Z.Attention in attention network for image super-resolution[J].arXiv:2104.09497,2021.
[14]MISRA D,NALAMADA T,ARASANIPALAIA U,et al.Rotate to attend:Convolutional triplet attention module[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision.2021:3139-3148.
[15]ZOU D,CAO Y,ZHOU D,et al.Gradient descent optimizesover-parameterized deep ReLU networks[J].Machine Lear-ning,2020,109(3):467-492.
[16]YANG X,XIE T,LIU L,et al.Image super-resolution reconstruction based on improved Dirac residual network[J].Multidimensional Systems and Signal Processing,2021,32(4):1065-1082.
[17]LAI W S,HUANG J B,AHUJA N,et al.Deep laplacian pyramid networks for fast and accurate super-resolution[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2017:624-632.
[18]WOO S,PARK J,LEE J Y,et al.Cbam:Convolutional block attention module[C]//Proceedings of the European Conference on Computer Vision(ECCV).2018:3-19.
[19]LYAQINI S,QUAFAFOU M,NACHAOUIM,et al.Supervised learning as an inverse problem based on non-smooth loss function[J].Knowledge and Information Systems,2020,62(8):3039-3058.
[20]TIMOFTE R,AGUSTSSON E,VAN GOOL L,et al.Ntire2017 challenge on single image super-resolution:Methods and results[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops.2017:114-125.
[21]CHEN Y,LIU L,PHONEVILAY V,et al.Image super-resolution reconstruction based on feature map attention mechanism[J].Applied Intelligence,2021,51(7):4367-4380.
[22]LIU Z,HUANG J,ZHU C,et al.Residual attention networkusing multi-channel dense connections for image super-resolution[J].Applied Intelligence,2021,51(1):85-99.
[23]LIM B,SON S,KIM H,et al.Enhanced deep residual networksfor single image super-resolution[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition Workshops.2017:136-144.
[24]LI F,CONG R,BAI H,et al.Deep interleaved network forimage super-resolution with asymmetric co-attention[J].arXiv:2004.11814,2020.
[25]DONG C,LOY C C,TANG X.Accelerating the super-resolution convolutional neural network[C]//European Conference on Computer Vision.Cham:Springer,2016:391-407.
[26]KIM J,LEE J K,LEE K M.Deeply-recursive convolutional network for image super-resolution[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.2016:1637-1645.
Viewed
Full text


Abstract

Cited
  Shared   
  Discussed   
No Suggested Reading articles found!
渝ICP备16013121号-4
地址：重庆市渝北区洪湖西路18号    邮编：401121  
电话：023-63500828（编辑部） 023-67039612（会议/专题） 023-67039623（财务/发票）
E-mail：jsjkx12@163.com
本系统由北京玛格泰克科技发展有限公司设计开发