[1]丁文超,张俊宝,阴庚雷.基于 CRNN 的 CSI 动作识别[J].计算机技术与发展,2021,31(06):7-12.[doi:10. 3969 / j. issn. 1673-629X. 2021. 06. 002]
　DING Wen-chao,ZHANG Jun-bao,YIN Geng-lei.CSI Action Recognition Based on CRNN[J].,2021,31(06):7-12.[doi:10. 3969 / j. issn. 1673-629X. 2021. 06. 002]
点击复制

基于 CRNN 的 CSI 动作识别()
分享到：

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

卷:
31
期数:
2021年06期
页码:
7-12
栏目:
人工智能
出版日期:
2021-06-10

文章信息/Info

Title:
CSI Action Recognition Based on CRNN
文章编号:
1673-629X(2021)06-0007-06
作者:
丁文超; 张俊宝; 阴庚雷
中原工学院 计算机学院,河南 郑州 450007
Author(s):
DING Wen-chao; ZHANG Jun-bao; YIN Geng-lei
School of Computer Science,Zhongyuan University of Technology,Zhengzhou 450007,China
关键词:
信道状态信息; CRNN; 动作识别; 自组织映射; 格拉姆角场; 链接时序分类
Keywords:
CSI; CRNN; action recognition; SOM; GASF; CTC
分类号:
TP391. 41
DOI:
10. 3969 / j. issn. 1673-629X. 2021. 06. 002
摘要:
随着 Wi-Fi 感知技术的发展,出现了大量使用 Wi-Fi 信道状态信息( channel state information,CSI)进行动作识别的应用。 然而大多数的方法在数据预处理和训练阶段都依赖于人工构建特征,构建过程耗时耗力并且需要专家的领域知识。 针对上述问题,提出一种基于 CRNN( convolutional recurrent neural network) 的 CSI 动作识别方法。 将不同手势的 CSI数据做低通滤波处理后,通过自组织映射( self organizing maps,SOM) 聚类的结果选择最佳子载波,并对该子载波上的 CSI数据进行扩增。 然后,使用格拉姆角求和场( Gramian angular summation fields,GASF) 方法将一维 CSI 数据转换成二维GASF 图像,作为 CNN、 LSTM 构 成 的 CRNN 网 络 的 输 入 数 据, 训 练 过 程 中 使 用 链 接 时 序 分 类 ( connectionist temporalclassification,CTC) 作为损失函数。 实验结果表明,该方法能在训练数据较少的情况下达到较高的识别精度,且无需手动构建特征。
Abstract:
With the development of Wi-Fi sensing technology,a large number of applications have emerged for motion recognition using channel state information ( CSI) from Wi - Fi. However, most methods rely on manual construction of features during the data pre -processing and training stages, the construction process is often time-consuming and labor - intensive and requires expert domain knowledge. Aiming at these problems,? we propose a CSI action recognition method based on convolutional recurrent neural network? ? ( CRNN) . After the CSI data of different gestures are processed by low-pass filter,the best subcarrier is selected by the result of self-organizing maps ( SOM) clustering,and the CSI data on the subcarrier is amplified. Then,one-dimensional CSI data is transformed into two-dimensional GASF images by using the method of gamma angular summation fields ( GASF) ,which is used as the input data of CRNN composed of CNN and LSTM. In the training process, the connectionist temporal classification ( CTC ) is used as the loss function. The experiment indicates that the proposed method can achieve high recognition accuracy with less training data,and there is no need to manually construct features.

相似文献/References:

[1]李耀辉,苏元茂. 一种基于直视路径识别的无线接入点定位方法[J].计算机技术与发展,2018,28(02):112.[doi:10．3969/j．issn．1673－629X．2018．02．025]
　LI Yao-hui,SU Yuan-mao.An Access Point Positioning Method Based on Line-of-sight Identification[J].,2018,28(06):112.[doi:10．3969/j．issn．1673－629X．2018．02．025]
[2]吴春香,张建明.无线感知网络中动作识别的滤波算法[J].计算机技术与发展,2018,28(05):86.[doi:10.3969/j.issn.1673－629X.2018.05.020]
　WU Chun-xiang,ZHANG Jian-ming.Filtering Algorithm for Motion Ｒecognition in Wireless Sensor Networks[J].,2018,28(06):86.[doi:10.3969/j.issn.1673－629X.2018.05.020]
[3]王拥军,马维华.一种基于 CSI 相位差的手势识别方法[J].计算机技术与发展,2020,30(04):14.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 003]
　WANG Yong-jun,MA Wei-hua.A Gesture Recognition Method Based on CSI Phase Difference[J].,2020,30(06):14.[doi:10. 3969 / j. issn. 1673-629X. 2020. 04. 003]
[4]刘 伟,程春玲,杜金楷,等.一种基于改进深度卷积模型的室内定位方法[J].计算机技术与发展,2022,32(08):155.[doi:10. 3969 / j. issn. 1673-629X. 2022. 08. 025]
　LIU Wei,CHENG Chun-ling,DU Jin-kai,et al.An Indoor Positioning Method Based on Improved Deep Convolution Mode[J].,2022,32(06):155.[doi:10. 3969 / j. issn. 1673-629X. 2022. 08. 025]
[5]朱德意,孙晴艺,董思凡,等.基于 OpenCV 的芯片 IMEI 码的检测与识别[J].计算机技术与发展,2022,32(08):174.[doi:10. 3969 / j. issn. 1673-629X. 2022. 08. 028]
　ZHU De-yi,SUN Qing-yi,DONG Si-fan,et al.Detection and Recognition of Chip IMEI Code Based on OpenCV[J].,2022,32(06):174.[doi:10. 3969 / j. issn. 1673-629X. 2022. 08. 028]
[6]马永连*,张登银.基于域自适应网络的跨场景摔倒检测算法研究[J].计算机技术与发展,2023,33(10):86.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 014]
　MA Yong-lian*,ZHANG Deng-yin.Cross-scene Fall Detection Algorithm Based on Domain Adaptive Network[J].,2023,33(06):86.[doi:10. 3969 / j. issn. 1673-629X. 2023. 10. 014]

更新日期/Last Update: 2021-06-10