Cargando…

Cross-Modality Person Re-Identification via Local Paired Graph Attention Network

Cross-modality person re-identification (ReID) aims at searching a pedestrian image of RGB modality from infrared (IR) pedestrian images and vice versa. Recently, some approaches have constructed a graph to learn the relevance of pedestrian images of distinct modalities to narrow the gap between IR...

Descripción completa

Detalles Bibliográficos
Autores principales: Zhou, Jianglin, Dong, Qing, Zhang, Zhong, Liu, Shuang, Durrani, Tariq S.
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2023
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10146823/
https://www.ncbi.nlm.nih.gov/pubmed/37112352
http://dx.doi.org/10.3390/s23084011
_version_ 1785034670854897664
author Zhou, Jianglin
Dong, Qing
Zhang, Zhong
Liu, Shuang
Durrani, Tariq S.
author_facet Zhou, Jianglin
Dong, Qing
Zhang, Zhong
Liu, Shuang
Durrani, Tariq S.
author_sort Zhou, Jianglin
collection PubMed
description Cross-modality person re-identification (ReID) aims at searching a pedestrian image of RGB modality from infrared (IR) pedestrian images and vice versa. Recently, some approaches have constructed a graph to learn the relevance of pedestrian images of distinct modalities to narrow the gap between IR modality and RGB modality, but they omit the correlation between IR image and RGB image pairs. In this paper, we propose a novel graph model called Local Paired Graph Attention Network (LPGAT). It uses the paired local features of pedestrian images from different modalities to build the nodes of the graph. For accurate propagation of information among the nodes of the graph, we propose a contextual attention coefficient that leverages distance information to regulate the process of updating the nodes of the graph. Furthermore, we put forward Cross-Center Contrastive Learning ([Formula: see text]) to constrain how far local features are from their heterogeneous centers, which is beneficial for learning the completed distance metric. We conduct experiments on the RegDB and SYSU-MM01 datasets to validate the feasibility of the proposed approach.
format Online
Article
Text
id pubmed-10146823
institution National Center for Biotechnology Information
language English
publishDate 2023
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-101468232023-04-29 Cross-Modality Person Re-Identification via Local Paired Graph Attention Network Zhou, Jianglin Dong, Qing Zhang, Zhong Liu, Shuang Durrani, Tariq S. Sensors (Basel) Article Cross-modality person re-identification (ReID) aims at searching a pedestrian image of RGB modality from infrared (IR) pedestrian images and vice versa. Recently, some approaches have constructed a graph to learn the relevance of pedestrian images of distinct modalities to narrow the gap between IR modality and RGB modality, but they omit the correlation between IR image and RGB image pairs. In this paper, we propose a novel graph model called Local Paired Graph Attention Network (LPGAT). It uses the paired local features of pedestrian images from different modalities to build the nodes of the graph. For accurate propagation of information among the nodes of the graph, we propose a contextual attention coefficient that leverages distance information to regulate the process of updating the nodes of the graph. Furthermore, we put forward Cross-Center Contrastive Learning ([Formula: see text]) to constrain how far local features are from their heterogeneous centers, which is beneficial for learning the completed distance metric. We conduct experiments on the RegDB and SYSU-MM01 datasets to validate the feasibility of the proposed approach. MDPI 2023-04-15 /pmc/articles/PMC10146823/ /pubmed/37112352 http://dx.doi.org/10.3390/s23084011 Text en © 2023 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Zhou, Jianglin
Dong, Qing
Zhang, Zhong
Liu, Shuang
Durrani, Tariq S.
Cross-Modality Person Re-Identification via Local Paired Graph Attention Network
title Cross-Modality Person Re-Identification via Local Paired Graph Attention Network
title_full Cross-Modality Person Re-Identification via Local Paired Graph Attention Network
title_fullStr Cross-Modality Person Re-Identification via Local Paired Graph Attention Network
title_full_unstemmed Cross-Modality Person Re-Identification via Local Paired Graph Attention Network
title_short Cross-Modality Person Re-Identification via Local Paired Graph Attention Network
title_sort cross-modality person re-identification via local paired graph attention network
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10146823/
https://www.ncbi.nlm.nih.gov/pubmed/37112352
http://dx.doi.org/10.3390/s23084011
work_keys_str_mv AT zhoujianglin crossmodalitypersonreidentificationvialocalpairedgraphattentionnetwork
AT dongqing crossmodalitypersonreidentificationvialocalpairedgraphattentionnetwork
AT zhangzhong crossmodalitypersonreidentificationvialocalpairedgraphattentionnetwork
AT liushuang crossmodalitypersonreidentificationvialocalpairedgraphattentionnetwork
AT durranitariqs crossmodalitypersonreidentificationvialocalpairedgraphattentionnetwork