A Mutiscale Residual Attention Network for Multitask Learning of Human Activity Using Radar Micro-Doppler Signatures
Short-range radar has become one of the latest sensor technologies for the Internet of Things (IoT), and it plays an increasingly vital role in IoT applications. As the essential task for various smart-sensing applications, radar-based human activity recognition and person identification have receiv...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2019-11-01
|
Series: | Remote Sensing |
Subjects: | |
Online Access: | https://www.mdpi.com/2072-4292/11/21/2584 |
id |
doaj-957eac711d8e41b6b8906bfa12e2d501 |
---|---|
record_format |
Article |
spelling |
doaj-957eac711d8e41b6b8906bfa12e2d5012020-11-25T01:35:59ZengMDPI AGRemote Sensing2072-42922019-11-011121258410.3390/rs11212584rs11212584A Mutiscale Residual Attention Network for Multitask Learning of Human Activity Using Radar Micro-Doppler SignaturesYuan He0Xinyu Li1Xiaojun Jing2Key Laboratory of Trustworthy Distributed Computing and Service (BUPT), Ministry of Education, Beijing University of Posts and Telecommunications, Beijing 100876, ChinaKey Laboratory of Trustworthy Distributed Computing and Service (BUPT), Ministry of Education, Beijing University of Posts and Telecommunications, Beijing 100876, ChinaKey Laboratory of Trustworthy Distributed Computing and Service (BUPT), Ministry of Education, Beijing University of Posts and Telecommunications, Beijing 100876, ChinaShort-range radar has become one of the latest sensor technologies for the Internet of Things (IoT), and it plays an increasingly vital role in IoT applications. As the essential task for various smart-sensing applications, radar-based human activity recognition and person identification have received more attention due to radar’s robustness to the environment and low power consumption. Activity recognition and person identification are generally treated as separate problems. However, designing different networks for these two tasks brings a high computational complexity and wastes of resources to some extent. Furthermore, there are some correlations in activity recognition and person identification tasks. In this work, we propose a multiscale residual attention network (<i>MRA-Net</i>) for joint activity recognition and person identification with radar micro-Doppler signatures. A fine-grained loss weight learning (FLWL) mechanism is presented for elaborating a multitask loss to optimize <i>MRA-Net</i>. In addition, we construct a new radar micro-Doppler dataset with dual labels of activity and identity. With the proposed model trained on this dataset, we demonstrate that our method achieves the state-of-the-art performance in both radar-based activity recognition and person identification tasks. The impact of the FLWL mechanism was further investigated, and ablation studies of the efficacy of each component in <i>MRA-Net</i> were also conducted.https://www.mdpi.com/2072-4292/11/21/2584smart sensinghuman activity recognitionperson identificationmultitask learningradar micro-doppler signatures |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Yuan He Xinyu Li Xiaojun Jing |
spellingShingle |
Yuan He Xinyu Li Xiaojun Jing A Mutiscale Residual Attention Network for Multitask Learning of Human Activity Using Radar Micro-Doppler Signatures Remote Sensing smart sensing human activity recognition person identification multitask learning radar micro-doppler signatures |
author_facet |
Yuan He Xinyu Li Xiaojun Jing |
author_sort |
Yuan He |
title |
A Mutiscale Residual Attention Network for Multitask Learning of Human Activity Using Radar Micro-Doppler Signatures |
title_short |
A Mutiscale Residual Attention Network for Multitask Learning of Human Activity Using Radar Micro-Doppler Signatures |
title_full |
A Mutiscale Residual Attention Network for Multitask Learning of Human Activity Using Radar Micro-Doppler Signatures |
title_fullStr |
A Mutiscale Residual Attention Network for Multitask Learning of Human Activity Using Radar Micro-Doppler Signatures |
title_full_unstemmed |
A Mutiscale Residual Attention Network for Multitask Learning of Human Activity Using Radar Micro-Doppler Signatures |
title_sort |
mutiscale residual attention network for multitask learning of human activity using radar micro-doppler signatures |
publisher |
MDPI AG |
series |
Remote Sensing |
issn |
2072-4292 |
publishDate |
2019-11-01 |
description |
Short-range radar has become one of the latest sensor technologies for the Internet of Things (IoT), and it plays an increasingly vital role in IoT applications. As the essential task for various smart-sensing applications, radar-based human activity recognition and person identification have received more attention due to radar’s robustness to the environment and low power consumption. Activity recognition and person identification are generally treated as separate problems. However, designing different networks for these two tasks brings a high computational complexity and wastes of resources to some extent. Furthermore, there are some correlations in activity recognition and person identification tasks. In this work, we propose a multiscale residual attention network (<i>MRA-Net</i>) for joint activity recognition and person identification with radar micro-Doppler signatures. A fine-grained loss weight learning (FLWL) mechanism is presented for elaborating a multitask loss to optimize <i>MRA-Net</i>. In addition, we construct a new radar micro-Doppler dataset with dual labels of activity and identity. With the proposed model trained on this dataset, we demonstrate that our method achieves the state-of-the-art performance in both radar-based activity recognition and person identification tasks. The impact of the FLWL mechanism was further investigated, and ablation studies of the efficacy of each component in <i>MRA-Net</i> were also conducted. |
topic |
smart sensing human activity recognition person identification multitask learning radar micro-doppler signatures |
url |
https://www.mdpi.com/2072-4292/11/21/2584 |
work_keys_str_mv |
AT yuanhe amutiscaleresidualattentionnetworkformultitasklearningofhumanactivityusingradarmicrodopplersignatures AT xinyuli amutiscaleresidualattentionnetworkformultitasklearningofhumanactivityusingradarmicrodopplersignatures AT xiaojunjing amutiscaleresidualattentionnetworkformultitasklearningofhumanactivityusingradarmicrodopplersignatures AT yuanhe mutiscaleresidualattentionnetworkformultitasklearningofhumanactivityusingradarmicrodopplersignatures AT xinyuli mutiscaleresidualattentionnetworkformultitasklearningofhumanactivityusingradarmicrodopplersignatures AT xiaojunjing mutiscaleresidualattentionnetworkformultitasklearningofhumanactivityusingradarmicrodopplersignatures |
_version_ |
1725064894926028800 |