基于神经网络的动作识别方法分析
|
摘要
提出基于神经网络的动作识别方法分析,可直接对时间序列数据进行处理,自动提取特征值,免去了人工提取特征值的繁琐过程。通过采集10个受试者的原始加速度数据,采用基于TensorFlow搭建的神经网络模型进行训练,从而对动作进行识别。实验结果表明:该系统能够快速有效的区分走、慢跑,上、下楼梯四种相似度较高的动作,平均识别率高达96.67%,最后和当前识别率高的两种传统机器学习方法相比较。
This paper proposes a neural network-based motion recognition method analysis,which can directly process time series data and automatically extract feature values,eliminating the cumbersome process of manually extracting feature values.By collecting the original acceleration data of 10 subjects,the neural network model based on TensorFlow is used for training to identify the action.The experimental results show that the system can quickly and effectively distinguish between four kinds of similar motions,such as walking and jogging,and the upper and lower stairs.The average recognition rate is as high as 96.67%.Finally,compared with the two traditional machine learning methods with high current recognition rate.
This paper proposes a neural network-based motion recognition method analysis,which can directly process time series data and automatically extract feature values,eliminating the cumbersome process of manually extracting feature values.By collecting the original acceleration data of 10 subjects,the neural network model based on TensorFlow is used for training to identify the action.The experimental results show that the system can quickly and effectively distinguish between four kinds of similar motions,such as walking and jogging,and the upper and lower stairs.The average recognition rate is as high as 96.67%.Finally,compared with the two traditional machine learning methods with high current recognition rate.
引文
[1]徐开先,马丽敏.传感器是国内物联网发展的瓶颈[J].仪表技术与传感器,2010(12):1-4.
[2]Chen L,Hoey J,Nugent C D,et al.Sensor-Based Activity Recognition[J].IEEE Transactions on Systems M an&Cybernetics Part C,2012,42(6):790-808.
[3]赵湛,韩璐,方震,等.基于可穿戴设备的日常压力状态评估研究[J].电子与信息学报,2017(11):2669-2676.
[4]Lara O D,Labrador M A.A Survey on Human Activity Recognition using Wearable Sensors[J].IEEE Communications Surveys&Tutorials,2013,15(3):1192-1209.
[5]Hsu Y L,Chung P C,Wang W H,et al.Gait and balance analysis for patients with Alzheimer's disease using an inertial-sensor-based wearable instrument.[J].IEEE Journal of Biomedical&Health Informatics,2014,18(6):1822.
[6]Gupta P,Dallas T.Feature selection and activity recognition system using a single triaxial accelerometer[J].IEEE Transactions on Biomedical Engineering,2014,61(6):1780.RONAO C A,CHO S B.Human activity recognition with smartphone sensors using deep learning neural networks[M].Oxford:Pergamon Press,Inc.,2016:235-244.
[7]ZENG M,LE T N,YU B,et al.Convolutional Neural Networks for Human Activity Recognition using Mobile Sensors:International Conference on Mobile Computing,Applications and Services,2015[C].ACM:Paris,France,2015.
[8]吴军,肖克聪.基于深度卷积神经网络的人体动作识别[J].华中科技大学学报(自然科学版),2016,44(s1).
[9]Lecun Y,Bengio Y,Hinton G.Deep learning[J].Nature,2015,521(7553):436.
[10]Peugh J L,Enders C K.Missing Data in Educational Research:A Review of Reporting Practices and Suggestions for Improvement[J].Review of Educational Research,2004,74(4):525-556.
[2]Chen L,Hoey J,Nugent C D,et al.Sensor-Based Activity Recognition[J].IEEE Transactions on Systems M an&Cybernetics Part C,2012,42(6):790-808.
[3]赵湛,韩璐,方震,等.基于可穿戴设备的日常压力状态评估研究[J].电子与信息学报,2017(11):2669-2676.
[4]Lara O D,Labrador M A.A Survey on Human Activity Recognition using Wearable Sensors[J].IEEE Communications Surveys&Tutorials,2013,15(3):1192-1209.
[5]Hsu Y L,Chung P C,Wang W H,et al.Gait and balance analysis for patients with Alzheimer's disease using an inertial-sensor-based wearable instrument.[J].IEEE Journal of Biomedical&Health Informatics,2014,18(6):1822.
[6]Gupta P,Dallas T.Feature selection and activity recognition system using a single triaxial accelerometer[J].IEEE Transactions on Biomedical Engineering,2014,61(6):1780.RONAO C A,CHO S B.Human activity recognition with smartphone sensors using deep learning neural networks[M].Oxford:Pergamon Press,Inc.,2016:235-244.
[7]ZENG M,LE T N,YU B,et al.Convolutional Neural Networks for Human Activity Recognition using Mobile Sensors:International Conference on Mobile Computing,Applications and Services,2015[C].ACM:Paris,France,2015.
[8]吴军,肖克聪.基于深度卷积神经网络的人体动作识别[J].华中科技大学学报(自然科学版),2016,44(s1).
[9]Lecun Y,Bengio Y,Hinton G.Deep learning[J].Nature,2015,521(7553):436.
[10]Peugh J L,Enders C K.Missing Data in Educational Research:A Review of Reporting Practices and Suggestions for Improvement[J].Review of Educational Research,2004,74(4):525-556.