{"title":"Applications of Deep Learning for Improved Recognition from Some High-Level Human Activities Using Sensors Data","authors":"Bhavantik Gondaliya, Anil Kumar Agrawal, Ankit Chouksey","doi":"10.1109/ASSIC55218.2022.10088361","DOIUrl":null,"url":null,"abstract":"More than half of the population of the world owns a smartphone, and many individuals are beginning to utilize smartwatches. Many real-world smartphones or smartwatch-based sensing applications are becoming available. To gain a better understanding of human behaviour, these applications recognize human activities using accelerometers and gyroscope sensors built into smartphones. In this research, we looked at the accelerometer and gyroscopes on both the smartphone and the smartwatch, as well as their combinations, to see which combination performs best for the underlying algorithms. This work demonstrates how to automatically extract discriminative features for activity recognition using Long Short Term Memory (LSTM) method, a deep learning approach. The results reported in this article show that using a smartwatch accelerometer and/or a combination of any two or four sensors can produce good results. However, we will endeavour to improve the accuracy of activity detection using raw sensor data.","PeriodicalId":441406,"journal":{"name":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-11-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 International Conference on Advancements in Smart, Secure and Intelligent Computing (ASSIC)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ASSIC55218.2022.10088361","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
More than half of the population of the world owns a smartphone, and many individuals are beginning to utilize smartwatches. Many real-world smartphones or smartwatch-based sensing applications are becoming available. To gain a better understanding of human behaviour, these applications recognize human activities using accelerometers and gyroscope sensors built into smartphones. In this research, we looked at the accelerometer and gyroscopes on both the smartphone and the smartwatch, as well as their combinations, to see which combination performs best for the underlying algorithms. This work demonstrates how to automatically extract discriminative features for activity recognition using Long Short Term Memory (LSTM) method, a deep learning approach. The results reported in this article show that using a smartwatch accelerometer and/or a combination of any two or four sensors can produce good results. However, we will endeavour to improve the accuracy of activity detection using raw sensor data.