No Cover Image

Journal article 396 views 45 downloads

Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition

Ruchita Mehta Orcid Logo, Sara Sharifzadeh Orcid Logo, Vasile Palade Orcid Logo, Bo Tan Orcid Logo, Alireza Daneshkhah Orcid Logo, Yordanka Karayaneva

Machine Learning and Knowledge Extraction, Volume: 5, Issue: 4, Pages: 1493 - 1518

Swansea University Author: Sara Sharifzadeh Orcid Logo

  • 64788.pdf

    PDF | Version of Record

    © 2023 by the authors. Licensee MDPI, Basel, Switzerland. Distributed under the terms of a Creative Commons Attribution 4.0 International License (CC BY 4.0).

    Download (4.19MB)

Check full text

DOI (Published version): 10.3390/make5040075

Abstract

Human capability to perform routine tasks declines with age and age-related problems. Remote human activity recognition (HAR) is beneficial for regular monitoring of the elderly population. This paper addresses the problem of the continuous detection of daily human activities using a mm-wave Doppler...

Full description

Published in: Machine Learning and Knowledge Extraction
ISSN: 2504-4990
Published: MDPI AG 2023
Online Access: Check full text

URI: https://cronfa.swan.ac.uk/Record/cronfa64788
first_indexed 2023-11-08T14:27:57Z
last_indexed 2024-11-25T14:14:44Z
id cronfa64788
recordtype SURis
fullrecord <?xml version="1.0"?><rfc1807><datestamp>2023-11-24T14:11:00.3282186</datestamp><bib-version>v2</bib-version><id>64788</id><entry>2023-10-20</entry><title>Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition</title><swanseaauthors><author><sid>a4e15f304398ecee3f28c7faec69c1b0</sid><ORCID>0000-0003-4621-2917</ORCID><firstname>Sara</firstname><surname>Sharifzadeh</surname><name>Sara Sharifzadeh</name><active>true</active><ethesisStudent>false</ethesisStudent></author></swanseaauthors><date>2023-10-20</date><deptcode>MACS</deptcode><abstract>Human capability to perform routine tasks declines with age and age-related problems. Remote human activity recognition (HAR) is beneficial for regular monitoring of the elderly population. This paper addresses the problem of the continuous detection of daily human activities using a mm-wave Doppler radar. In this study, two strategies have been employed: the first method uses un-equalized series of activities, whereas the second method utilizes a gradient-based strategy for equalization of the series of activities. The dynamic time warping (DTW) algorithm and Long Short-term Memory (LSTM) techniques have been implemented for the classification of un-equalized and equalized series of activities, respectively. The input for DTW was provided using three strategies. The first approach uses the pixel-level data of frames (UnSup-PLevel). In the other two strategies, a convolutional variational autoencoder (CVAE) is used to extract Un-Supervised Encoded features (UnSup-EnLevel) and Supervised Encoded features (Sup-EnLevel) from the series of Doppler frames. The second approach for equalized data series involves the application of four distinct feature extraction methods: i.e., convolutional neural networks (CNN), supervised and unsupervised CVAE, and principal component Analysis (PCA). The extracted features were considered as an input to the LSTM. This paper presents a comparative analysis of a novel supervised feature extraction pipeline, employing Sup-ENLevel-DTW and Sup-EnLevel-LSTM, against several state-of-the-art unsupervised methods, including UnSUp-EnLevel-DTW, UnSup-EnLevel-LSTM, CNN-LSTM, and PCA-LSTM. The results demonstrate the superiority of the Sup-EnLevel-LSTM strategy. However, the UnSup-PLevel strategy worked surprisingly well without using annotations and frame equalization.</abstract><type>Journal Article</type><journal>Machine Learning and Knowledge Extraction</journal><volume>5</volume><journalNumber>4</journalNumber><paginationStart>1493</paginationStart><paginationEnd>1518</paginationEnd><publisher>MDPI AG</publisher><placeOfPublication/><isbnPrint/><isbnElectronic/><issnPrint/><issnElectronic>2504-4990</issnElectronic><keywords>Human activity recognition (HAR), dynamic time warping (DTW), convolutional variational autoencoder (CVAE), mm-wave radar sensor, deep neural networks (DNNs)</keywords><publishedDay>14</publishedDay><publishedMonth>10</publishedMonth><publishedYear>2023</publishedYear><publishedDate>2023-10-14</publishedDate><doi>10.3390/make5040075</doi><url>http://dx.doi.org/10.3390/make5040075</url><notes/><college>COLLEGE NANME</college><department>Mathematics and Computer Science School</department><CollegeCode>COLLEGE CODE</CollegeCode><DepartmentCode>MACS</DepartmentCode><institution>Swansea University</institution><apcterm>Not Required</apcterm><funders>Coventry University</funders><projectreference/><lastEdited>2023-11-24T14:11:00.3282186</lastEdited><Created>2023-10-20T10:40:10.5911227</Created><path><level id="1">Faculty of Science and Engineering</level><level id="2">School of Mathematics and Computer Science - Computer Science</level></path><authors><author><firstname>Ruchita</firstname><surname>Mehta</surname><orcid>0009-0008-8164-8250</orcid><order>1</order></author><author><firstname>Sara</firstname><surname>Sharifzadeh</surname><orcid>0000-0003-4621-2917</orcid><order>2</order></author><author><firstname>Vasile</firstname><surname>Palade</surname><orcid>0000-0002-6768-8394</orcid><order>3</order></author><author><firstname>Bo</firstname><surname>Tan</surname><orcid>0000-0002-6855-6270</orcid><order>4</order></author><author><firstname>Alireza</firstname><surname>Daneshkhah</surname><orcid>0000-0001-7751-4307</orcid><order>5</order></author><author><firstname>Yordanka</firstname><surname>Karayaneva</surname><order>6</order></author></authors><documents><document><filename>64788__28975__476c24a4c3bc4fca91c2c26f76f06c34.pdf</filename><originalFilename>64788.pdf</originalFilename><uploaded>2023-11-08T14:30:23.9827336</uploaded><type>Output</type><contentLength>4388329</contentLength><contentType>application/pdf</contentType><version>Version of Record</version><cronfaStatus>true</cronfaStatus><documentNotes>&#xA9; 2023 by the authors. Licensee MDPI, Basel, Switzerland. Distributed under the terms of a Creative Commons Attribution 4.0 International License (CC BY 4.0).</documentNotes><copyrightCorrect>true</copyrightCorrect><language>eng</language><licence>https://creativecommons.org/licenses/by/4.0/</licence></document></documents><OutputDurs/></rfc1807>
spelling 2023-11-24T14:11:00.3282186 v2 64788 2023-10-20 Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition a4e15f304398ecee3f28c7faec69c1b0 0000-0003-4621-2917 Sara Sharifzadeh Sara Sharifzadeh true false 2023-10-20 MACS Human capability to perform routine tasks declines with age and age-related problems. Remote human activity recognition (HAR) is beneficial for regular monitoring of the elderly population. This paper addresses the problem of the continuous detection of daily human activities using a mm-wave Doppler radar. In this study, two strategies have been employed: the first method uses un-equalized series of activities, whereas the second method utilizes a gradient-based strategy for equalization of the series of activities. The dynamic time warping (DTW) algorithm and Long Short-term Memory (LSTM) techniques have been implemented for the classification of un-equalized and equalized series of activities, respectively. The input for DTW was provided using three strategies. The first approach uses the pixel-level data of frames (UnSup-PLevel). In the other two strategies, a convolutional variational autoencoder (CVAE) is used to extract Un-Supervised Encoded features (UnSup-EnLevel) and Supervised Encoded features (Sup-EnLevel) from the series of Doppler frames. The second approach for equalized data series involves the application of four distinct feature extraction methods: i.e., convolutional neural networks (CNN), supervised and unsupervised CVAE, and principal component Analysis (PCA). The extracted features were considered as an input to the LSTM. This paper presents a comparative analysis of a novel supervised feature extraction pipeline, employing Sup-ENLevel-DTW and Sup-EnLevel-LSTM, against several state-of-the-art unsupervised methods, including UnSUp-EnLevel-DTW, UnSup-EnLevel-LSTM, CNN-LSTM, and PCA-LSTM. The results demonstrate the superiority of the Sup-EnLevel-LSTM strategy. However, the UnSup-PLevel strategy worked surprisingly well without using annotations and frame equalization. Journal Article Machine Learning and Knowledge Extraction 5 4 1493 1518 MDPI AG 2504-4990 Human activity recognition (HAR), dynamic time warping (DTW), convolutional variational autoencoder (CVAE), mm-wave radar sensor, deep neural networks (DNNs) 14 10 2023 2023-10-14 10.3390/make5040075 http://dx.doi.org/10.3390/make5040075 COLLEGE NANME Mathematics and Computer Science School COLLEGE CODE MACS Swansea University Not Required Coventry University 2023-11-24T14:11:00.3282186 2023-10-20T10:40:10.5911227 Faculty of Science and Engineering School of Mathematics and Computer Science - Computer Science Ruchita Mehta 0009-0008-8164-8250 1 Sara Sharifzadeh 0000-0003-4621-2917 2 Vasile Palade 0000-0002-6768-8394 3 Bo Tan 0000-0002-6855-6270 4 Alireza Daneshkhah 0000-0001-7751-4307 5 Yordanka Karayaneva 6 64788__28975__476c24a4c3bc4fca91c2c26f76f06c34.pdf 64788.pdf 2023-11-08T14:30:23.9827336 Output 4388329 application/pdf Version of Record true © 2023 by the authors. Licensee MDPI, Basel, Switzerland. Distributed under the terms of a Creative Commons Attribution 4.0 International License (CC BY 4.0). true eng https://creativecommons.org/licenses/by/4.0/
title Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition
spellingShingle Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition
Sara Sharifzadeh
title_short Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition
title_full Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition
title_fullStr Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition
title_full_unstemmed Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition
title_sort Deep Learning Techniques for Radar-Based Continuous Human Activity Recognition
author_id_str_mv a4e15f304398ecee3f28c7faec69c1b0
author_id_fullname_str_mv a4e15f304398ecee3f28c7faec69c1b0_***_Sara Sharifzadeh
author Sara Sharifzadeh
author2 Ruchita Mehta
Sara Sharifzadeh
Vasile Palade
Bo Tan
Alireza Daneshkhah
Yordanka Karayaneva
format Journal article
container_title Machine Learning and Knowledge Extraction
container_volume 5
container_issue 4
container_start_page 1493
publishDate 2023
institution Swansea University
issn 2504-4990
doi_str_mv 10.3390/make5040075
publisher MDPI AG
college_str Faculty of Science and Engineering
hierarchytype
hierarchy_top_id facultyofscienceandengineering
hierarchy_top_title Faculty of Science and Engineering
hierarchy_parent_id facultyofscienceandengineering
hierarchy_parent_title Faculty of Science and Engineering
department_str School of Mathematics and Computer Science - Computer Science{{{_:::_}}}Faculty of Science and Engineering{{{_:::_}}}School of Mathematics and Computer Science - Computer Science
url http://dx.doi.org/10.3390/make5040075
document_store_str 1
active_str 0
description Human capability to perform routine tasks declines with age and age-related problems. Remote human activity recognition (HAR) is beneficial for regular monitoring of the elderly population. This paper addresses the problem of the continuous detection of daily human activities using a mm-wave Doppler radar. In this study, two strategies have been employed: the first method uses un-equalized series of activities, whereas the second method utilizes a gradient-based strategy for equalization of the series of activities. The dynamic time warping (DTW) algorithm and Long Short-term Memory (LSTM) techniques have been implemented for the classification of un-equalized and equalized series of activities, respectively. The input for DTW was provided using three strategies. The first approach uses the pixel-level data of frames (UnSup-PLevel). In the other two strategies, a convolutional variational autoencoder (CVAE) is used to extract Un-Supervised Encoded features (UnSup-EnLevel) and Supervised Encoded features (Sup-EnLevel) from the series of Doppler frames. The second approach for equalized data series involves the application of four distinct feature extraction methods: i.e., convolutional neural networks (CNN), supervised and unsupervised CVAE, and principal component Analysis (PCA). The extracted features were considered as an input to the LSTM. This paper presents a comparative analysis of a novel supervised feature extraction pipeline, employing Sup-ENLevel-DTW and Sup-EnLevel-LSTM, against several state-of-the-art unsupervised methods, including UnSUp-EnLevel-DTW, UnSup-EnLevel-LSTM, CNN-LSTM, and PCA-LSTM. The results demonstrate the superiority of the Sup-EnLevel-LSTM strategy. However, the UnSup-PLevel strategy worked surprisingly well without using annotations and frame equalization.
published_date 2023-10-14T02:37:38Z
_version_ 1821280733971349504
score 11.047306

Similar Items