Image Classification Algorithm Based on Lightweight Feature Fusion Convolutional Network

doi:10.19678/j.issn.1000-3428.0059815

Abstract

Abstract: The traditional Convolutional Neural Networks (CNN) suffer from single convolutional kernels,complex network structure and redundant parameters.To address the problem,a lightweight CNN named MS-FNet is designed for feature fusion.The fusion module employs a multi-branch structure to increase the width of the CNN,and different sizes of convolutional kernels to process the input feature map,which improves the ability of the network to extract different features in the same layer.And the redundant features are removed after each convolution by using BN,ReLU,etc.Convolutional layers are used to replace the traditional fully connected layer,which not only accelerates the training speed of model but also alleviates overfitting problems caused by too many parameters.The experimental results show that MS-FNet greatly reduces the number of network parameters and the error rate.

Key words: Deep Learning(DL), Convolutional Neural Network(CNN), feature extraction, feature fusion, image classification

摘要： 传统卷积神经网络存在卷积核单一、网络结构复杂和参数冗余的问题。提出一种轻量级特征融合卷积神经网络MS-FNet。在融合模块中采用多路结构以增加卷积神经网络的宽度，通过不同尺寸的卷积核对输入特征图进行处理，提高网络在同一层中提取不同特征的能力，并在每次卷积后采用批归一化、ReLU等方法去除冗余特征。此外，使用卷积层代替传统的全连接层，从而加快模型的训练速度，缓解因参数过多造成的过拟合现象。实验结果表明，MS-FNet可在降低错误率的同时，有效减少网络参数量。

关键词: 深度学习, 卷积神经网络, 特征提取, 特征融合, 图像分类

CLC Number:

TP391.4

CHEN Xinhua, QIAN Xuezhong, SONG Wei. Image Classification Algorithm Based on Lightweight Feature Fusion Convolutional Network[J]. Computer Engineering, 2021, 47(11): 268-275.

陈鑫华, 钱雪忠, 宋威. 基于轻量级特征融合卷积网络的图像分类算法[J]. 计算机工程, 2021, 47(11): 268-275.

/ / Recommend / Download Citations

URL: http://www.ecice06.com/EN/10.19678/j.issn.1000-3428.0059815

http://www.ecice06.com/EN/Y2021/V47/I11/268

Figures/Tables 15

References

[1] 张顺,龚怡宏,王进军.深度卷积神经网络的发展及其在计算机视觉领域的应用[J].计算机学报,2019,42(3):453-482. ZHANG S,GONG Y H,WANG J J.The development of deep convolution neural network and its application of computer vision[J].Chinese Journal of Computers,2019,42(3):453-482.(in Chinese)
[2] HU Y,SUN S,LI J,et al.A novel channel pruning method for deep neural network compression[EB/OL].[2020-09-15].http://export.arxiv.org/pdf/1805.11394.
[3] 靳丽蕾,杨文柱,王思乐,等.一种用于卷积神经网络压缩的混合剪枝方法[J].小型微型计算机系统,2018,39(12):2596-2601. JIN L L,YANG W Z,WANG S L,et al.Mixed pruning method for convolutional neural network compression[J].Journal of Chinese Mini-Micro Computer Systems,2018,39(12):2596-2601.(in Chinese)
[4] IANDOLA F,HAN S,MOSKEWICZ M W,et al.SqueezeNet:AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size[EB/OL].[2020-09-20].https://arxiv.org/abs/1602.07360v1.
[5] ZHANG X,ZHOU X,LIN M,et al.ShuffleNet:an extremely efficient convolutional neural network for mobile devices[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2018:6848-6856.
[6] MEHTA S,RASTEGARI M,CASPI A,et al.ESPNet:efficient spatial pyramid of dilated convolutions for semantic segmentation[C]//Proceedings of European Conference on Computer Vision.Berlin,Germany:Springer,2018:552-568.
[7] SZEGEDY C,LIU W,JIA Y,et al.Going deeper with convolutions[C]//Proceedings of 2015 IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2015:1-9.
[8] HE K,ZHANG X,REN S,et al.Deep residual learning for image recognition[C]//Proceedings of 2016 IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2016:770-778.
[9] HUANG G,LIU Z,MAATEN L V D,et al.Densely connected convolutional networks[C]//Proceedings of 2017 IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2017:2261-2269.
[10] ZHANG Z,WANG X,JUNG C.DCSR:dilated convolutions for single image super-resolution[J].IEEE Transactions on Image Processing,2019,28(4):1625-1635.
[11] SZEGEDY C,VANHOUCKE V,IOFFE S,et al.Rethinking the inception architecture for computer vision[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C.,USA:IEEE Press,2016:2818-2826.
[12] ARORA S,BHASKARA A,GE R,et al.Provable bounds for learning some deep representations[EB/OL].[2020-09-21].https://arxiv.org/abs/1310.6343v1.
[13] SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[EB/OL].[2020-09-18].https://arxiv.org/abs/1409.1556.
[14] LIN M,CHEN Q,YAN S,et al.Network in network[EB/OL].[2020-09-22].https://arxiv.org/abs/1312.4400.
[15] 高云龙,吴川,朱明.基于改进卷积神经网络的短文本分类模型[J].吉林大学学报(理学版),2020,58(4):923-930. GAO Y L,WU C,ZHU M.Short text classification model based on improved convolutional neural network[J].Journal of Jilin University(Science Edition),2020,58(4):923-930.(in Chinese)
[16] ROY A G,NAVAB N,WACHINGER C.Recalibrating fully convolutional networks with spatial and channel "squeeze and excitation" blocks[J].IEEE Transactions on Medical Imaging,2019,38(2):540-549.
[17] LI Y,WANG N,SHI J,et al.Adaptive batch normalization for practical domain adaptation[J].Pattern Recognition,2018,80(4):109-117.
[18] ABADI M,BARHAM P,CHEN J,et al.TensorFlow:a system for large-scale machine learning[C]//Proceedings of Symposium on Operating Systems Design and Implementation.New York,USA:ACM Press,2016:265-283.
[19] LEE C,XIE S,GALLAGHER P W,et al.Deeply-supervised nets[C]//Proceedings of International Conference on Artificial Intelligence and Statistics.New York,USA:ACM Press,2015:562-570.
[20] 杨萌林,张文生.分类激活图增强的图像分类算法[J].计算机科学与探索,2020,14(1):149-158. YANG M L,ZHANG W S.Image classification algorithm based on classification activation map enhancement[J].Journal of Frontiers of Computer Science and Technology,2020,14(1):149-158.(in Chinese)
[21] ZAGORUYKO S,KOMODAKIS N.Wide residual networks[EB/OL].[2020-09-20].https://arxiv.org/abs/1605.07146v4.
[22] LARSSON G,MAIRE M,SHAKHNAROVICH G,et al.FractalNet:ultra-deep neural networks without residuals[EB/OL].[2020-09-21].https://arxiv.org/abs/1605.07648v1.
[23] WU Y.Deep convolutional neural network based on densely connected squeeze-and-excitation blocks[J].AIP Advances,2019,9(6):1-10.
[24] HOU S,LIU X,WANG Z.Dualnet:learn complementary features for image recognition[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C.,USA:IEEE Press,2017:502-510.
[25] VELICKOVIC P,WANG D,LANEY N D,et al.X-CNN:cross-modal convolutional neural networks for sparse datasets[C]//Proceedings of IEEE Symposium Series on Computational Intelligence.Washington D.C.,USA:IEEE Press,2016:1-8.
[26] KILINC O,UYSAL I.GAR:an efficient and scalable graph-based activity regularization for semi-supervised learning[J].Neurocomputing,2018,296(6):46-54.
[27] 付晓,沈远彤,李宏伟,等.基于半监督编码生成对抗网络的图像分类模型[J].自动化学报,2020,46(3):531-539. FU X,SHEN Y T,LI H W,et al.A semi-supervised encoder generative adversarial networks model for image classification[J].Acta Automatica Sinica,2020,46(3):531-539.(in Chinese)

Please choose a citation manager

Content to export