Jointly adaptive cross-resolution person re-identification on super-resolution

Abstract Cross-resolution Person Re-identification (ReID) faces the significant challenge of large resolution variance across different camera views in real surveillance systems. Most approaches based on super-resolution (SR) excessively rely on the SR images, which may lead to the loss of low-resol...

Full description

Saved in:
Bibliographic Details
Main Authors: Caihong Yuan, Zhijie Guan, Yuanchen Xu, Xiaopan Chen, Xiaoke Zhu, Wenjuan Liang
Format: Article
Language:English
Published: Springer 2025-04-01
Series:Complex & Intelligent Systems
Subjects:
Online Access:https://doi.org/10.1007/s40747-025-01881-1
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849733919365136384
author Caihong Yuan
Zhijie Guan
Yuanchen Xu
Xiaopan Chen
Xiaoke Zhu
Wenjuan Liang
author_facet Caihong Yuan
Zhijie Guan
Yuanchen Xu
Xiaopan Chen
Xiaoke Zhu
Wenjuan Liang
author_sort Caihong Yuan
collection DOAJ
description Abstract Cross-resolution Person Re-identification (ReID) faces the significant challenge of large resolution variance across different camera views in real surveillance systems. Most approaches based on super-resolution (SR) excessively rely on the SR images, which may lead to the loss of low-resolution (LR) information. Meanwhile, the region-agnostic SR could pose interference to ReID. For this, we propose a jointly adaptive cross-resolution ReID framework that consists of a region-aware person super-resolution (RAPSR) and a resolution adaptive ReID (RAReID). RAPSR is equipped with spatial attention for enhancing crucial spatial regions in low-resolution (LR) images. RAReID extracts complementary features from LR and high-resolution (HR) images simultaneously and obtains more discriminative pedestrian representations through cascaded resolution adaptive feature fusion modules. Finally, by the joint training of RAPSR and RAReID, a greater ReID accuracy could be achieved. Extensive experiments demonstrate state-of-the-art performances on three derived and a native multi-resolution datasets.
format Article
id doaj-art-df3bffbfd4d3411aaf2bf2ad2bc0a019
institution DOAJ
issn 2199-4536
2198-6053
language English
publishDate 2025-04-01
publisher Springer
record_format Article
series Complex & Intelligent Systems
spelling doaj-art-df3bffbfd4d3411aaf2bf2ad2bc0a0192025-08-20T03:07:55ZengSpringerComplex & Intelligent Systems2199-45362198-60532025-04-0111611110.1007/s40747-025-01881-1Jointly adaptive cross-resolution person re-identification on super-resolutionCaihong Yuan0Zhijie Guan1Yuanchen Xu2Xiaopan Chen3Xiaoke Zhu4Wenjuan Liang5Henan Province Spatial Information Processing Engineering Technology Research Center Henan Engineering Research Center of Intelligent Technology and Application, Henan UniversityHenan Province Spatial Information Processing Engineering Technology Research Center Henan Engineering Research Center of Intelligent Technology and Application, Henan UniversityHenan Province Spatial Information Processing Engineering Technology Research Center Henan Engineering Research Center of Intelligent Technology and Application, Henan UniversityHenan Province Spatial Information Processing Engineering Technology Research Center Henan Engineering Research Center of Intelligent Technology and Application, Henan UniversityHenan Province Spatial Information Processing Engineering Technology Research Center Henan Engineering Research Center of Intelligent Technology and Application, Henan UniversityHenan Province Spatial Information Processing Engineering Technology Research Center Henan Engineering Research Center of Intelligent Technology and Application, Henan UniversityAbstract Cross-resolution Person Re-identification (ReID) faces the significant challenge of large resolution variance across different camera views in real surveillance systems. Most approaches based on super-resolution (SR) excessively rely on the SR images, which may lead to the loss of low-resolution (LR) information. Meanwhile, the region-agnostic SR could pose interference to ReID. For this, we propose a jointly adaptive cross-resolution ReID framework that consists of a region-aware person super-resolution (RAPSR) and a resolution adaptive ReID (RAReID). RAPSR is equipped with spatial attention for enhancing crucial spatial regions in low-resolution (LR) images. RAReID extracts complementary features from LR and high-resolution (HR) images simultaneously and obtains more discriminative pedestrian representations through cascaded resolution adaptive feature fusion modules. Finally, by the joint training of RAPSR and RAReID, a greater ReID accuracy could be achieved. Extensive experiments demonstrate state-of-the-art performances on three derived and a native multi-resolution datasets.https://doi.org/10.1007/s40747-025-01881-1Cross-resolutionPerson re-identificationRegion-aware super-resolutionResolution adaptive
spellingShingle Caihong Yuan
Zhijie Guan
Yuanchen Xu
Xiaopan Chen
Xiaoke Zhu
Wenjuan Liang
Jointly adaptive cross-resolution person re-identification on super-resolution
Complex & Intelligent Systems
Cross-resolution
Person re-identification
Region-aware super-resolution
Resolution adaptive
title Jointly adaptive cross-resolution person re-identification on super-resolution
title_full Jointly adaptive cross-resolution person re-identification on super-resolution
title_fullStr Jointly adaptive cross-resolution person re-identification on super-resolution
title_full_unstemmed Jointly adaptive cross-resolution person re-identification on super-resolution
title_short Jointly adaptive cross-resolution person re-identification on super-resolution
title_sort jointly adaptive cross resolution person re identification on super resolution
topic Cross-resolution
Person re-identification
Region-aware super-resolution
Resolution adaptive
url https://doi.org/10.1007/s40747-025-01881-1
work_keys_str_mv AT caihongyuan jointlyadaptivecrossresolutionpersonreidentificationonsuperresolution
AT zhijieguan jointlyadaptivecrossresolutionpersonreidentificationonsuperresolution
AT yuanchenxu jointlyadaptivecrossresolutionpersonreidentificationonsuperresolution
AT xiaopanchen jointlyadaptivecrossresolutionpersonreidentificationonsuperresolution
AT xiaokezhu jointlyadaptivecrossresolutionpersonreidentificationonsuperresolution
AT wenjuanliang jointlyadaptivecrossresolutionpersonreidentificationonsuperresolution