Survey of Cross-Modal Person Re-Identification from a Mathematical Perspective
Person re-identification (Re-ID) aims to retrieve a particular pedestrian’s identification from a surveillance system consisting of non-overlapping cameras. In recent years, researchers have begun to focus on open-world person Re-ID tasks based on non-ideal situations. One of the most representative...
Main Authors: | , , |
---|---|
Format: | Article |
Language: | English |
Published: |
MDPI AG
2023-01-01
|
Series: | Mathematics |
Subjects: | |
Online Access: | https://www.mdpi.com/2227-7390/11/3/654 |
_version_ | 1797623827218300928 |
---|---|
author | Minghui Liu Yafei Zhang Huafeng Li |
author_facet | Minghui Liu Yafei Zhang Huafeng Li |
author_sort | Minghui Liu |
collection | DOAJ |
description | Person re-identification (Re-ID) aims to retrieve a particular pedestrian’s identification from a surveillance system consisting of non-overlapping cameras. In recent years, researchers have begun to focus on open-world person Re-ID tasks based on non-ideal situations. One of the most representative of these is cross-modal person Re-ID, which aims to match probe data with target data from different modalities. According to the modalities of probe and target data, we divided cross-modal person Re-ID into visible–infrared, visible–depth, visible–sketch, and visible–text person Re-ID. In cross-modal person Re-ID, the most challenging problem is the modal gap. According to the different methods of narrowing the modal gap, we classified the existing works into picture-based style conversion methods, feature-based modality-invariant embedding mapping methods, and modality-unrelated auxiliary information mining methods. In addition, by generalizing the aforementioned works, we find that although deep-learning-based models perform well, the black-box-like learning process makes these models less interpretable and generalized. Therefore, we attempted to interpret different cross-modal person Re-ID models from a mathematical perspective. Through the above work, we attempt to compensate for the lack of mathematical interpretation of models in previous person Re-ID reviews and hope that our work will bring new inspiration to researchers. |
first_indexed | 2024-03-11T09:34:16Z |
format | Article |
id | doaj.art-543f0468dcf845e9a9b7421ab92e62d5 |
institution | Directory Open Access Journal |
issn | 2227-7390 |
language | English |
last_indexed | 2024-03-11T09:34:16Z |
publishDate | 2023-01-01 |
publisher | MDPI AG |
record_format | Article |
series | Mathematics |
spelling | doaj.art-543f0468dcf845e9a9b7421ab92e62d52023-11-16T17:22:42ZengMDPI AGMathematics2227-73902023-01-0111365410.3390/math11030654Survey of Cross-Modal Person Re-Identification from a Mathematical PerspectiveMinghui Liu0Yafei Zhang1Huafeng Li2Faculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, ChinaFaculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, ChinaFaculty of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650500, ChinaPerson re-identification (Re-ID) aims to retrieve a particular pedestrian’s identification from a surveillance system consisting of non-overlapping cameras. In recent years, researchers have begun to focus on open-world person Re-ID tasks based on non-ideal situations. One of the most representative of these is cross-modal person Re-ID, which aims to match probe data with target data from different modalities. According to the modalities of probe and target data, we divided cross-modal person Re-ID into visible–infrared, visible–depth, visible–sketch, and visible–text person Re-ID. In cross-modal person Re-ID, the most challenging problem is the modal gap. According to the different methods of narrowing the modal gap, we classified the existing works into picture-based style conversion methods, feature-based modality-invariant embedding mapping methods, and modality-unrelated auxiliary information mining methods. In addition, by generalizing the aforementioned works, we find that although deep-learning-based models perform well, the black-box-like learning process makes these models less interpretable and generalized. Therefore, we attempted to interpret different cross-modal person Re-ID models from a mathematical perspective. Through the above work, we attempt to compensate for the lack of mathematical interpretation of models in previous person Re-ID reviews and hope that our work will bring new inspiration to researchers.https://www.mdpi.com/2227-7390/11/3/654cross-modal person re-identificationreviewmathematical perspective |
spellingShingle | Minghui Liu Yafei Zhang Huafeng Li Survey of Cross-Modal Person Re-Identification from a Mathematical Perspective Mathematics cross-modal person re-identification review mathematical perspective |
title | Survey of Cross-Modal Person Re-Identification from a Mathematical Perspective |
title_full | Survey of Cross-Modal Person Re-Identification from a Mathematical Perspective |
title_fullStr | Survey of Cross-Modal Person Re-Identification from a Mathematical Perspective |
title_full_unstemmed | Survey of Cross-Modal Person Re-Identification from a Mathematical Perspective |
title_short | Survey of Cross-Modal Person Re-Identification from a Mathematical Perspective |
title_sort | survey of cross modal person re identification from a mathematical perspective |
topic | cross-modal person re-identification review mathematical perspective |
url | https://www.mdpi.com/2227-7390/11/3/654 |
work_keys_str_mv | AT minghuiliu surveyofcrossmodalpersonreidentificationfromamathematicalperspective AT yafeizhang surveyofcrossmodalpersonreidentificationfromamathematicalperspective AT huafengli surveyofcrossmodalpersonreidentificationfromamathematicalperspective |