Pedestrian Pose Recognition Based on Frequency-Modulated Continuous-Wave Radar with Meta-Learning.

Shi, J; Zhang, Q; Shi, Q; Chu, L; Braun, R

Pedestrian Pose Recognition Based on Frequency-Modulated Continuous-Wave Radar with Meta-Learning.

Shi, J Zhang, Q Shi, Q Chu, L Braun, R

Permalink

Publisher:: MDPI
Publication Type:: Journal Article
Citation:: Sensors (Basel), 2024, 24, (9), pp. 2932
Issue Date:: 2024-05-05

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download Published versionAdobe PDF (8.69 MB)

View on publisher's site

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Shi, J
dc.contributor.author	Zhang, Q
dc.contributor.author	Shi, Q
dc.contributor.author	Chu, L
dc.contributor.author	Braun, R https://orcid.org/0000-0002-4153-8769
dc.date.accessioned	2024-08-02T02:45:25Z
dc.date.available	2024-05-02
dc.date.available	2024-08-02T02:45:25Z
dc.date.issued	2024-05-05
dc.identifier.citation	Sensors (Basel), 2024, 24, (9), pp. 2932
dc.identifier.issn	1424-8220
dc.identifier.issn	1424-8220
dc.identifier.uri	http://hdl.handle.net/10453/180001
dc.description.abstract	With the continuous advancement of autonomous driving and monitoring technologies, there is increasing attention on non-intrusive target monitoring and recognition. This paper proposes an ArcFace SE-attention model-agnostic meta-learning approach (AS-MAML) by integrating attention mechanisms into residual networks for pedestrian gait recognition using frequency-modulated continuous-wave (FMCW) millimeter-wave radar through meta-learning. We enhance the feature extraction capability of the base network using channel attention mechanisms and integrate the additive angular margin loss function (ArcFace loss) into the inner loop of MAML to constrain inner loop optimization and improve radar discrimination. Then, this network is used to classify small-sample micro-Doppler images obtained from millimeter-wave radar as the data source for pose recognition. Experimental tests were conducted on pose estimation and image classification tasks. The results demonstrate significant detection and recognition performance, with an accuracy of 94.5%, accompanied by a 95% confidence interval. Additionally, on the open-source dataset DIAT-μRadHAR, which is specially processed to increase classification difficulty, the network achieves a classification accuracy of 85.9%.
dc.format	Electronic
dc.language	eng
dc.publisher	MDPI
dc.relation.ispartof	Sensors (Basel)
dc.relation.isbasedon	10.3390/s24092932
dc.rights	info:eu-repo/semantics/openAccess
dc.subject	0301 Analytical Chemistry, 0502 Environmental Science and Management, 0602 Ecology, 0805 Distributed Computing, 0906 Electrical and Electronic Engineering
dc.subject.classification	Analytical Chemistry
dc.subject.classification	3103 Ecology
dc.subject.classification	4008 Electrical engineering
dc.subject.classification	4009 Electronics, sensors and digital hardware
dc.subject.classification	4104 Environmental management
dc.subject.classification	4606 Distributed computing and systems software
dc.subject.mesh	Humans
dc.subject.mesh	Pedestrians
dc.subject.mesh	Radar
dc.subject.mesh	Algorithms
dc.subject.mesh	Gait
dc.subject.mesh	Pattern Recognition, Automated
dc.subject.mesh	Machine Learning
dc.subject.mesh	Humans
dc.subject.mesh	Gait
dc.subject.mesh	Algorithms
dc.subject.mesh	Radar
dc.subject.mesh	Pattern Recognition, Automated
dc.subject.mesh	Machine Learning
dc.subject.mesh	Pedestrians
dc.subject.mesh	Humans
dc.subject.mesh	Pedestrians
dc.subject.mesh	Radar
dc.subject.mesh	Algorithms
dc.subject.mesh	Gait
dc.subject.mesh	Pattern Recognition, Automated
dc.subject.mesh	Machine Learning
dc.title	Pedestrian Pose Recognition Based on Frequency-Modulated Continuous-Wave Radar with Meta-Learning.
dc.type	Journal Article
utslib.citation.volume	24
utslib.location.activity	Switzerland
utslib.for	0301 Analytical Chemistry
utslib.for	0502 Environmental Science and Management
utslib.for	0602 Ecology
utslib.for	0805 Distributed Computing
utslib.for	0906 Electrical and Electronic Engineering
pubs.organisational-group	University of Technology Sydney
pubs.organisational-group	University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	University of Technology Sydney/Strength - CRIN - Realtime Information Networks
pubs.organisational-group	University of Technology Sydney/Strength - GBDTC - Global Big Data Technologies
pubs.organisational-group	University of Technology Sydney/All Manual Groups
pubs.organisational-group	University of Technology Sydney/All Manual Groups/Global Big Data Technologies Research Centre (GBDTC)
utslib.copyright.status	open_access	*
dc.rights.license	This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0). To view a copy of this license, visit https://creativecommons.org/licenses/by/4.0/
dc.date.updated	2024-08-02T02:45:18Z
pubs.issue	9
pubs.publication-status	Published online
pubs.volume	24
utslib.citation.issue	9

Abstract:

With the continuous advancement of autonomous driving and monitoring technologies, there is increasing attention on non-intrusive target monitoring and recognition. This paper proposes an ArcFace SE-attention model-agnostic meta-learning approach (AS-MAML) by integrating attention mechanisms into residual networks for pedestrian gait recognition using frequency-modulated continuous-wave (FMCW) millimeter-wave radar through meta-learning. We enhance the feature extraction capability of the base network using channel attention mechanisms and integrate the additive angular margin loss function (ArcFace loss) into the inner loop of MAML to constrain inner loop optimization and improve radar discrimination. Then, this network is used to classify small-sample micro-Doppler images obtained from millimeter-wave radar as the data source for pose recognition. Experimental tests were conducted on pose estimation and image classification tasks. The results demonstrate significant detection and recognition performance, with an accuracy of 94.5%, accompanied by a 95% confidence interval. Additionally, on the open-source dataset DIAT-μRadHAR, which is specially processed to increase classification difficulty, the network achieves a classification accuracy of 85.9%.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/180001