|本期目录/Table of Contents|

[1]田秋红,孙文轩,章立早,等.基于改进GhostNet的轻量级手势图像识别方法[J].浙江理工大学学报,2023,49-50(自科三):300-309.
　TIAN Qiuhong,SUN Wenxuan,ZHANG Lizao,et al.Lightweight gesture image recognition method  based on improved GhostNet[J].Journal of Zhejiang Sci-Tech University,2023,49-50(自科三):300-309.
点击复制

基于改进GhostNet的轻量级手势图像识别方法()
分享到：

浙江理工大学学报[ISSN:1673-3851/CN:33-1338/TS]

卷:
第49-50卷
期数:
2023年自科第三期
页码:
300-309
栏目:
出版日期:
2023-05-31

文章信息/Info

Title:
Lightweight gesture image recognition method  based on improved GhostNet
文章编号:
1673-3851 (2023) 05-0300-10
作者:
田秋红; 孙文轩; 章立早; 施之翔; 潘豪; 吴佳璐
浙江理工大学计算机科学与技术学院，杭州 310018
Author(s):
TIAN Qiuhong;  SUN Wenxuan;  ZHANG Lizao;  SHI Zhixiang;  PAN Hao;  WU Jialu
School of Computer Science and Technology, Zhejiang  Sci-Tech University, Hangzhou 310018, China
关键词:
手势图像识别; 卷积神经网络; 轻量级模型; 注意力机制; 激活函数
分类号:
TP181
文献标志码:
A
摘要:
卷积神经网络应用于复杂背景的手势图像识别时，存在深层模型参数量大、计算成本高、轻量级模型准确率低等问题，针对这些问题提出了一种基于改进GhostNet的轻量级手势图像识别方法。首先，在Ghost模块中添加通道混洗操作，建立CS Ghost模块以提取手势图像中的手势特征；然后，选用SMU（Smoothing maximum unit）激活函数优化模型在反向传播中的学习能力；最后，使用注意力机制中的轻量级通道注意力模块ECA去除特征中的噪声信息。该方法在ASL和NUS Ⅱ数据集上的实验平均准确率分别为98.82%和99.36%；在OUHANDS数据集上的实验平均准确率为97.98%，参数量为1.2 Mi，FLOPs为0.29 Gi。实验结果表明该方法参数量小，计算成本低，可有效提高手势图像识别的准确率。

参考文献/References:

［ 1 ］ Jiang D, Zheng Z J, Li G F, et al. Gesture recognition based on binocular vision ［ J ］ . Cluster Computing, 2019, 22(6): 13261 - 13271.

［2］王银, 陈云龙, 孙前来. 复杂背景下的手势识别［J］. 中国图象图形学报, 2021, 26(4):815-827.

［3］陈影柔, 田秋红, 杨慧敏, 等. 基于多特征加权融合的静态手势识别［J］. 计算机系统应用, 2021, 30(2):20-27.

［4］Tian Q H, Bao J X, Yang H M, et al. Improving arm segmentation in sign language recognition systems using image processing［J］. Technology and Health Care: Official Journal of the European Society for Engineering and Medicine, 2021, 29(3): 527-540.

［5］Sadeddine K, Chelali F Z, Djeradi R, et al. Recognition of user-dependent and independent static hand gestures: Application to sign language［J］. Journal of Visual Communication and Image Representation, 2021, 79: 103193.

［6］杨述斌, 潘伟, 蒋宗霖. 基于HOG特征与手部多特征信息融合的静态手势识别［J］. 自动化与仪表, 2020, 35(8):47-51.

［7］Pardasani A, Sharma A K, Banerjee S, et al. Enhancing the ability to communicate by synthesizing american sign language using image recognition in a chatbot for differently abled［C］∥2018 7th International Conference on Reliability, Infocom Technologies and Optimization(Trends and Future Directions). Noida, India: IEEE, 2018: 529-532.

［8］Khotimah W N, Suciati N, Benedict I. Indonesian sign language recognition by using the static and dynamic features［C］∥2018 International Seminar on Intelligent Technology and Its Applications (ISITIA). Bali, Indonesia: IEEE, 2018: 293-298.

［9］Kwolek B, Baczynski W, Sako S. Recognition of JSL fingerspelling using deep convolutional neural networks［J］. Neurocomputing, 2021, 456: 586-598.

［10］Xie B, He X Y, Li Y. RGB-D static gesture recognition based on convolutional neural network［J］. The Journal of Engineering, 2018, 2018(16): 1515-1520.

undefined

相似文献/References:

[1]李斯凡,高法钦.基于卷积神经网络的手写数字识别[J].浙江理工大学学报,2017,37-38(自科3):438.
　LI Sifan,GAO Faqin.Handwritten Numeral Recognition Based on Convolution Neural Network[J].Journal of Zhejiang Sci-Tech University,2017,37-38(自科三):438.
[2]张玮,张华熊.基于卷积神经网络的纺织面料主成分分类[J].浙江理工大学学报,2019,41-42(自科一):1.
　ZHANG Wei,ZHANG Huaxiong.Classification of main components of textile fabrics based on convolutional neural network[J].Journal of Zhejiang Sci-Tech University,2019,41-42(自科三):1.
[3]邓远远,沈炜.基于注意力反馈机制的深度图像标注模型[J].浙江理工大学学报,2019,41-42(自科二):208.
　DENG Yuanyuan,SHEN Wei.Depth image caption model based on  attention feedback mechanism[J].Journal of Zhejiang Sci-Tech University,2019,41-42(自科三):208.
[4]邓远远,沈炜.基于注意力反馈机制的深度图像标注模型[J].浙江理工大学学报,2019,41-42(自科二):208.
　DENG Yuanyuan,SHEN Wei.Depth image caption model based on attention feedback mechanism[J].Journal of Zhejiang Sci-Tech University,2019,41-42(自科三):208.
[5]陈巧红,董雯,孙麒,等.基于混合神经网络的单文档自动文摘模型[J].浙江理工大学学报,2019,41-42(自科四):489.
　CHEN Qiaohong,DONG Wen,SUN Qi,et al.Single document automatic summarization model based on hybrid neural network[J].Journal of Zhejiang Sci-Tech University,2019,41-42(自科三):489.
[6]陈巧红,王磊,孙麒,等.基于混合神经网络的中文短文本分类模型[J].浙江理工大学学报,2019,41-42(自科四):509.
　CHEN Qiaohong,WANG Lei,SUN Qi,et al.Chinese short text classification model based on hybrid neural network[J].Journal of Zhejiang Sci-Tech University,2019,41-42(自科三):509.
[7]程诚,任佳.基于自适应卷积核的改进CNN数值型数据分类算法[J].浙江理工大学学报,2019,41-42(自科五):657.
　CHENG Cheng,REN Jia.Improved CNN classification algorithm based on adaptive convolution kernel for numerical data[J].Journal of Zhejiang Sci-Tech University,2019,41-42(自科三):657.
[8]祝鹏烜,黄体仁,李旭.MSAG-TransNet：肺部CT图像中新冠肺炎感染新型冠状病毒感染区域的分割模型[J].浙江理工大学学报,2023,49-50(自科六):734.
　ZHU Pengxuan,HUANG Tiren,LI Xu.MSAG-TransNet: Segmentation model of COVID-19 infected areas in lung CT images[J].Journal of Zhejiang Sci-Tech University,2023,49-50(自科三):734.

备注/Memo

备注/Memo:
收稿日期： 2022-10-31
网络出版日期：2023-01-16
基金项目： 国家自然科学基金项目(51405448);基于实践课程的学生创新能力培育方法研究(11120033312202)；浙江省教育厅一般科研项目(Y202250600);浙江省大学生科技创新活动计划大学生科技创新项目（2022R406A014）
作者简介： 田秋红(1976-)，女，辽宁兴城人，教授，博士，主要从事机器学习、模式识别和图像处理与识别方面的研究。
更新日期/Last Update: 2023-09-08