基于多尺度分割的图像识别残差网络研究

doi:10.19678/j.issn.1000-3428.0061392

摘要/Abstract

摘要： 深度卷积神经网络能够解决复杂的计算机视觉问题，被广泛应用于图像识别任务中。在基于深度卷积神经网络的图像识别过程中，增加网络的深度和宽度能够产生丰富的特征信息，使用多尺度分割方法能够有效减少冗余的特征信息。然而，增加网络的深度和进行多尺度分割都会影响识别速度。如何在保证精度的同时提高识别速度，成为设计高效网络的关键问题。通过增加网络宽度的方法对ResNet残差网络进行改进，在保证精度的基础上提升识别速度。使用ResNet-D中的残差结构并减少网络长度，得到长度只有7层的残差网络，同时对HS-ResNet中的多尺度分割方法进行优化，只保留最后一次连接合并操作，得到图像识别残差网络SSRNet。在CIFAR 10和CIFAR 100数据集上的实验结果显示，SSRNet速度最高较ResNet网络提升7倍多，同时错误率最高下降8.81%，表明缩短网络长度可大幅加快图像识别速度，同时结合多尺度分割方法能够有效提升识别精度。

关键词: 多尺度分割, 残差网络, 卷积神经网络, 图像识别, 图像分类

Abstract: The emergence of deep Convolutional Neural Network(CNN) has contributed significantly to solving complex computer vision problems, and they have been widely used in image recognition tasks.Designing an efficient network has become the key to improving the performance of deep CNN.In the process of image recognition based on a deep CNN, increasing the depth and width of the network can produce rich feature information, whereas the use of a multi-scale segmentation method can effectively reduce redundant feature information.However, increasing the depth of the network in multi-scale segmentation affects the recognition speed.Thus, improving recognition speed while ensuring accuracy has become an important goal in designing efficient networks.To solve this problem, the network width is increased using ResNet to improve the recognition speed and ensure accuracy.Using the residual structure in ResNet-D and reducing the network length, a residual network with only seven layers is obtained.Concurrently, the multi-scale segmentation method in HS-ResNet is optimized, and only the last connection and merging operation are retained to obtain SSRNet.The experimental results on the CIFAR 10 and CIFAR 100 datasets show that the maximum speed of SSRNet is more than seven times higher than that of ResNet, and the error rate can be reduced by 8.81%.This demonstrates that shortening the length of the network can significantly accelerate the speed of image recognition, whereby the recognition accuracy is effectively improved in combination with the multi-scale segmentation method.

Key words: multiscale split, residual network, Convolutional Neural Network(CNN), image recognition, image classification

中图分类号:

TP391

袁单飞, 陈慈发, 董方敏. 基于多尺度分割的图像识别残差网络研究[J]. 计算机工程, 2022, 48(5): 258-262,271.

YUAN Danfei, CHEN Cifa, DONG Fangmin. Research on Residual Network of Image Recognition Based on Multiscale Split[J]. Computer Engineering, 2022, 48(5): 258-262,271.

https://www.ecice06.com/CN/Y2022/V48/I5/258

图/表 7

20220806174816

20220806174820

20220806174824

20220806174828

20220806174831

20220806174835

20220806174838

参考文献

[1] KRIZHEVSKY A, SUTSKEVER I, HINTON G E.ImageNet classification with deep convolutional neural networks[J].Communications of the ACM, 2017, 60(6):84-90.
[2] SIMONYAN K, ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[EB/OL].[2021-06-11].https://arxiv.org/abs/1409.1556.
[3] HE K M, ZHANG X Y, REN S Q, et al.Deep residual learning for image recognition[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2016:770-778.
[4] ZAGORUYKO S, KOMODAKIS N.Wide residual networks[C]//Proceedings of British Machine Vision Conference.[S.l.]:BMVC, 2016:1-12.
[5] GAO S H, CHENG M M, ZHAO K, et al.Res2Net:a new multi-scale backbone architecture[J].IEEE Transactions on Pattern Analysis and Machine Intelligence, 2021, 43(2):652-662.
[6] YUAN P C, LIN S F, CUI C, et al.HS-ResNet:hierarchical-split block on convolutional neural network[EB/OL].[2021-06-11].https://arxiv.org/abs/2010.07621.
[7] HE T, ZHANG Z, ZHANG H, et al.Bag of tricks for image classification with convolutional neural networks[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2019:558-567.
[8] HAN D, KIM J, KIM J.Deep pyramidal residual networks[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA:IEEE Press, 2017:6307-6315.
[9] 任长娥, 袁超, 孙彦丽, 等.宽度学习系统研究进展[J].计算机应用研究, 2021, 38(8):2258-2267. REN C G, YUAN C, SUN Y L, et al.Research of broad learning system[J].Application Research of Computers, 2021, 38(8):2258-2267.(in Chinese)
[10] 黎玲利, 孟令兵, 李金宝.多尺度特征提取和多级别特征融合的显著性目标检测方法[J].工程科学与技术, 2021, 53(1):170-177. LI L L, MENG L B, LI J B.Salient object detection based on multi-scale feature extraction and multi-level feature fusion[J].Advanced Engineering Sciences, 2021, 53(1):170-177.(in Chinese)
[11] 官申珂, 林晓, 郑晓妹, 等.结合超像素分割的多尺度特征融合图像语义分割算法[J].图学学报, 2021, 42(3):406-413. GUAN S K, LIN X, ZHENG X M, et al.A semantic segmentation algorithm using multi-scale feature fusion with combination of superpixel segmentation[J].Journal of Graphics, 2021, 42(3):406-413.(in Chinese)
[12] 任欢, 王旭光.注意力机制综述[J].计算机应用, 2021, 41(S1):1-6. REN H, WANG X G.Review of attention mechanism[J].Journal of Computer Applications, 2021, 41(S1):1-6.(in Chinese)
[13] 赵升, 赵黎.基于双向特征金字塔和深度学习的图像识别方法[J].哈尔滨理工大学学报, 2021, 26(2):44-50. ZHAO S, ZHAO L.On image recognition using bidirectional feature pyramid and deep neural network[J].Journal of Harbin University of Science and Technology, 2021(2):44-50.(in Chinese)
[14] 朱旭东, 熊贇.基于多层次注意力和图模型的图像多标签分类研究[J/OL].计算机工程:1-8[2021-06-11].DOI:10.19678/j.issn.1000-3428.0061072. ZHU X D, XIONG Y.Multi-label image classification method based on multi scale attention and graph model[J/OL].Computer Engineering:1-8[2021-06-11].DOI:10.19678/j.issn.1000-3428.0061072.(in Chinese)
[15] 吴旭, 刘翔, 赵静文.一种轻量级多尺度融合的图像篡改检测算法[J].计算机工程, 2022, 48(2):224-229, 236. WU X, LIU X, ZHAO J W.A lightweight multiscale fusion algorithm for image tampering detection[J].Computer Engineering, 2022, 48(2):224-229, 236.(in Chinese)
[16] 王柳程, 欧阳城添, 梁文.基于改进特征金字塔网络的人体姿态估计[J].计算机工程, 2021, 47(8):251-259, 270. WANG L C, OUYANG C T, LIANG W.Human pose estimation based on improved pyramid feature network[J].Computer Engineering, 2021, 47(8):251-259, 270.(in Chinese)
[17] LIN T Y, DOLLÁR P, GIRSHICK R, et al.Feature pyramid networks for object detection[C]//Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.Washington D.C., USA, USA:IEEE Press, 2017:936-944.
[18] HAN K, WANG Y H, TIAN Q, et al.GhostNet:more features from cheap operations[C]//Proceedings of IEEE/CVF Conference on Computer Vision and Pattern Recognition.Washington D.C., USA, USA:IEEE Press, 2020:1577-1586.
[19] KRIZHEVSKY A, HINTON G.Learning multiple layers of features from tiny images[J].Handbook of Systemic Autoimmune Diseases, 2009, 1(4):1-5.
[20] LUO W G, LI Y J, URTASUN R, et al.Understanding the effective receptive field in deep convolutional neural networks[EB/OL].[2021-06-11].https://arxiv.org/pdf/1701.04128v1.pdf.
[21] CAO X.A practical theory for designing very deep convo-lutional neural networks[EB/OL].[2021-06-11].http://pdfs.semanticscholar.org/7922/2fad9f671be142bd7e42cd785a2cb06a1d30.pdf.
[22] HE K M, ZHANG X Y, REN S Q, et al.Delving deep into rectifiers:surpassing human-level performance on ImageNet classification[C]//Proceedings of IEEE International Conference on Computer Vision.Washington D.C., USA, USA:IEEE Press, 2015:1026-1034.

选择文件类型/文献管理软件名称

选择包含的内容