Deep Metric Learning for Scalable Gait-Based Person Re-Identification Using Force Platform Data
Walking gait data acquired with force platforms may be used for person re-identification (re-ID) in various authentication, surveillance, and forensics applications. Current force platform-based re-ID systems classify a fixed set of identities (IDs), which presents a problem when IDs are added or re...
Published in: | Sensors |
---|---|
Main Authors: | , , , , , , |
Format: | Text |
Language: | English |
Published: |
Multidisciplinary Digital Publishing Institute
2023
|
Subjects: | |
Online Access: | https://doi.org/10.3390/s23073392 |
id |
ftmdpi:oai:mdpi.com:/1424-8220/23/7/3392/ |
---|---|
record_format |
openpolar |
spelling |
ftmdpi:oai:mdpi.com:/1424-8220/23/7/3392/ 2023-08-20T04:06:09+02:00 Deep Metric Learning for Scalable Gait-Based Person Re-Identification Using Force Platform Data Kayne A. Duncanson Simon Thwaites David Booth Gary Hanly William S. P. Robertson Ehsan Abbasnejad Dominic Thewlis 2023-03-23 application/pdf https://doi.org/10.3390/s23073392 EN eng Multidisciplinary Digital Publishing Institute Physical Sensors https://dx.doi.org/10.3390/s23073392 https://creativecommons.org/licenses/by/4.0/ Sensors; Volume 23; Issue 7; Pages: 3392 gait recognition biometric ground reaction force deep learning zero-shot re-ID center of pressure time series gait analysis force plate classification Text 2023 ftmdpi https://doi.org/10.3390/s23073392 2023-08-01T09:24:06Z Walking gait data acquired with force platforms may be used for person re-identification (re-ID) in various authentication, surveillance, and forensics applications. Current force platform-based re-ID systems classify a fixed set of identities (IDs), which presents a problem when IDs are added or removed from the database. We formulated force platform-based re-ID as a deep metric learning (DML) task, whereby a deep neural network learns a feature representation that can be compared between inputs using a distance metric. The force platform dataset used in this study is one of the largest and the most comprehensive of its kind, containing 193 IDs with significant variations in clothing, footwear, walking speed, and time between trials. Several DML model architectures were evaluated in a challenging setting where none of the IDs were seen during training (i.e., zero-shot re-ID) and there was only one prior sample per ID to compare with each query sample. The best architecture was 85% accurate in this setting, though an analysis of changes in walking speed and footwear between measurement instances revealed that accuracy was 28% higher on same-speed, same-footwear comparisons, compared to cross-speed, cross-footwear comparisons. These results demonstrate the potential of DML algorithms for zero-shot re-ID using force platform data, and highlight challenging cases. Text DML MDPI Open Access Publishing Sensors 23 7 3392 |
institution |
Open Polar |
collection |
MDPI Open Access Publishing |
op_collection_id |
ftmdpi |
language |
English |
topic |
gait recognition biometric ground reaction force deep learning zero-shot re-ID center of pressure time series gait analysis force plate classification |
spellingShingle |
gait recognition biometric ground reaction force deep learning zero-shot re-ID center of pressure time series gait analysis force plate classification Kayne A. Duncanson Simon Thwaites David Booth Gary Hanly William S. P. Robertson Ehsan Abbasnejad Dominic Thewlis Deep Metric Learning for Scalable Gait-Based Person Re-Identification Using Force Platform Data |
topic_facet |
gait recognition biometric ground reaction force deep learning zero-shot re-ID center of pressure time series gait analysis force plate classification |
description |
Walking gait data acquired with force platforms may be used for person re-identification (re-ID) in various authentication, surveillance, and forensics applications. Current force platform-based re-ID systems classify a fixed set of identities (IDs), which presents a problem when IDs are added or removed from the database. We formulated force platform-based re-ID as a deep metric learning (DML) task, whereby a deep neural network learns a feature representation that can be compared between inputs using a distance metric. The force platform dataset used in this study is one of the largest and the most comprehensive of its kind, containing 193 IDs with significant variations in clothing, footwear, walking speed, and time between trials. Several DML model architectures were evaluated in a challenging setting where none of the IDs were seen during training (i.e., zero-shot re-ID) and there was only one prior sample per ID to compare with each query sample. The best architecture was 85% accurate in this setting, though an analysis of changes in walking speed and footwear between measurement instances revealed that accuracy was 28% higher on same-speed, same-footwear comparisons, compared to cross-speed, cross-footwear comparisons. These results demonstrate the potential of DML algorithms for zero-shot re-ID using force platform data, and highlight challenging cases. |
format |
Text |
author |
Kayne A. Duncanson Simon Thwaites David Booth Gary Hanly William S. P. Robertson Ehsan Abbasnejad Dominic Thewlis |
author_facet |
Kayne A. Duncanson Simon Thwaites David Booth Gary Hanly William S. P. Robertson Ehsan Abbasnejad Dominic Thewlis |
author_sort |
Kayne A. Duncanson |
title |
Deep Metric Learning for Scalable Gait-Based Person Re-Identification Using Force Platform Data |
title_short |
Deep Metric Learning for Scalable Gait-Based Person Re-Identification Using Force Platform Data |
title_full |
Deep Metric Learning for Scalable Gait-Based Person Re-Identification Using Force Platform Data |
title_fullStr |
Deep Metric Learning for Scalable Gait-Based Person Re-Identification Using Force Platform Data |
title_full_unstemmed |
Deep Metric Learning for Scalable Gait-Based Person Re-Identification Using Force Platform Data |
title_sort |
deep metric learning for scalable gait-based person re-identification using force platform data |
publisher |
Multidisciplinary Digital Publishing Institute |
publishDate |
2023 |
url |
https://doi.org/10.3390/s23073392 |
genre |
DML |
genre_facet |
DML |
op_source |
Sensors; Volume 23; Issue 7; Pages: 3392 |
op_relation |
Physical Sensors https://dx.doi.org/10.3390/s23073392 |
op_rights |
https://creativecommons.org/licenses/by/4.0/ |
op_doi |
https://doi.org/10.3390/s23073392 |
container_title |
Sensors |
container_volume |
23 |
container_issue |
7 |
container_start_page |
3392 |
_version_ |
1774717082136477696 |