[1]张国有,高 希.融合注意力的轻量型垃圾分类研究[J].计算机技术与发展,2023,33(03):49-56.[doi:10. 3969 / j. issn. 1673-629X. 2023. 03. 008]
　ZHANG Guo-you,GAO Xi.Research on Lightweight Garbage Sorting with Fusion Attention[J].,2023,33(03):49-56.[doi:10. 3969 / j. issn. 1673-629X. 2023. 03. 008]
点击复制

融合注意力的轻量型垃圾分类研究()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
33
期数:
2023年03期
页码:
49-56
栏目:
媒体计算
出版日期:
2023-03-10

文章信息/Info

Title:
Research on Lightweight Garbage Sorting with Fusion Attention
文章编号:
1673-629X(2023)03-0049-08
作者:
张国有; 高 希
太原科技大学 计算机科学与技术学院,山西 太原 030024
Author(s):
ZHANG Guo-you; GAO Xi
School of Computer Science and Technology,Taiyuan University of Science and Technology,Taiyuan 030024,China
关键词:
卷积神经网络; 垃圾分类; 轻量化; 特征融合; 迁移学习
Keywords:
convolutional neural network; garbage classification; lightweight; feature fusion; transfer learning
分类号:
TP391
DOI:
10. 3969 / j. issn. 1673-629X. 2023. 03. 008
摘要:
针对轻量化网络在图像分类任务中无法直接部署在小型计算机,如:树莓派开发板,且存在检测速度慢、对硬件资源要求较高的问题,提出了一种基于 ShuffleNet 的改进算法。 首先,将传统卷积替换为最大公约数分组卷积,以减少网络所需的参数量和计算量,降低网络对于硬件算计资源的需求;其次,通过引入 SE 模块,融合通道注意力信息,提升网络在分类任务的检测精度;最后,针对多级分类的垃圾分类问题,分别连接不同节点数量的全连接层以及对标准类别和细分类别分别引入不同的损失和准确率权重,提升网络在多级垃圾分类任务的能力。 将改进的 ShuffleNet 网络和 ShuffleNet 网络中具有不同网络层数的版本,进行准确率和速率方面的对比。 实验结果表明,改进后的网络在基本的分类任务准确率达到 80% 以上,且其中的 0. 5 版本能够直接部署在树莓派开发板,平均单张图像处理时间 1. 28 s,降低了网络对于硬件资源的需求。
Abstract:
Aiming at the problem that lightweight networks cannot be directly deployed on small computers such as Raspberry Pidevelopment boards in image classification tasks, and there are problems of slow detection speed and high hardware resourcerequirements,an improved algorithm based on ShuffleNet is proposed. First, the traditional convolution is replaced by the greatestcommon divisor grouped convolution to reduce the amount of parameters and computation required by the network, and reduce thedemand of network for hardware computing resources. In addition,by introducing the SE module,the channel attention information isfused to improve the detection accuracy of the network in the classification task. Finally,for the garbage classification problem of multi-level classification,the fully connected layers with different numbers of nodes are respectively connected,and different loss and accuracyweights are introduced for the standard category and subdivision category respectively,　so as to improve the ability of the network in themulti-level garbage classification task. The improved ShuffleNet network and the version with different network layers 　in the ShuffleNetnetwork are compared in terms of accuracy and speed. The experimental results show that the accuracy rate of the improved network inthe basic classification task reaches more than 80% ,and the 0. 5 version can be directly deployed on the Raspberry Pi development board.The average single image processing time is 1. 28 s,which reduces the hardware resources demand of network.

相似文献/References:

[1]崔凤焦.表情识别算法研究进展与性能比较[J].计算机技术与发展,2018,28(02):145.[doi:10．3969/j．issn．1673－629X．2018．02．031]
　CUI Feng-jiao.Ｒesearch and Performance Comparison of Facial Expression Ｒecognition Algorithm[J].,2018,28(03):145.[doi:10．3969/j．issn．1673－629X．2018．02．031]
[2]张丹丹,李雷. 基于PCANet-RF的人脸检测系统[J].计算机技术与发展,2016,26(02):31.
　ZHANG Dan-dan,LI Lei. Face Detection System Based on PCANet-RF[J].,2016,26(03):31.
[3]陈强锐,谢世朋.基于深度学习的肺部肿瘤检测方法[J].计算机技术与发展,2018,28(04):201.[doi:10.3969/ j. issn.1673-629X.2018.04.043]
　CHEN Qiang-rui,XIE Shi-peng.Lung Cancer Detection Method Based on Deep Learning[J].,2018,28(03):201.[doi:10.3969/ j. issn.1673-629X.2018.04.043]
[4]郭子琰,舒心,刘常燕,等.基于ReLU 函数的卷积神经网络的花卉识别算法[J].计算机技术与发展,2018,28(05):154.[doi:10．3969/j．issn．1673－629X．2018．05．035]
　GUO Ziyan,SHU Xin,LIU Changyan,et al.A Recognition Algorithm of Flower Based on Convolution Neural Network with ReLU Function[J].,2018,28(03):154.[doi:10．3969/j．issn．1673－629X．2018．05．035]
[5]缪宇杰,吴智钧,宫 婧.基于3D 卷积的视频错帧筛选方法[J].计算机技术与发展,2018,28(05):179.[doi:10.3969/ j. issn.1673-629X.2018.05.040]
　MIAO Yu-jie,WU Zhi-jun,GONG Jing.A Wrong Temporal-order Frames Identification Method Based on 3D Convolution[J].,2018,28(03):179.[doi:10.3969/ j. issn.1673-629X.2018.05.040]
[6]吴玉枝,吴志红,熊运余.基于卷积神经网络的小样本车辆检测与识别[J].计算机技术与发展,2018,28(06):1.[doi:10.3969/ j. issn.1673-629X.2018.06.001]
　WU Yu-zhi,WU Zhi-hong,XIONG Yun-yu.Vehicle Detection and Recognition of a Few Samples Based on Convolutional Neural Network[J].,2018,28(03):1.[doi:10.3969/ j. issn.1673-629X.2018.06.001]
[7]李相桥,李晨,田丽华,等.卷积神经网络并行训练的优化研究[J].计算机技术与发展,2018,28(08):12.[doi:10.3969/ j. issn.1673-629X.2018.08.003]
　LI Xiang-qiao,LI Chen,TIAN Li-hua,et al.Research on Optimization of Parallel Training for Convolution Neural Network[J].,2018,28(03):12.[doi:10.3969/ j. issn.1673-629X.2018.08.003]
[8]邓宗平,赵启军,陈虎. 基于深度学习的人脸姿态分类方法[J].计算机技术与发展,2016,26(07):11.
　DEND Zong-ping,ZHAO Qi-jun,CHEN Hu. Face Pose Classification Method Based on Deep Learning[J].,2016,26(03):11.
[9]河海大学 计算机与信息学院,江苏 南京 0098.卷积网络的无监督特征提取对人脸识别的研究[J].计算机技术与发展,2018,28(06):17.[doi:10.3969/ j. issn.1673-629X.2018.06.004]
　DU Bai-sheng.Research on Unsupervised Feature Extraction Based on Convolutional Neural Network for Face Recognition[J].,2018,28(03):17.[doi:10.3969/ j. issn.1673-629X.2018.06.004]
[10]高翔,陈志,岳文静,等.基于视频场景深度学习的人物语义识别模型[J].计算机技术与发展,2018,28(06):53.[doi:10.3969/ j. issn.1673-629X.2018.06.012]
　GAO Xiang,CHEN Zhi,YUE Wen-jing,et al.Human Semantic Recognition Model Based on Video Scene Deep Learning[J].,2018,28(03):53.[doi:10.3969/ j. issn.1673-629X.2018.06.012]

更新日期/Last Update: 2023-03-10