A Lightweight Model Based on Co-segmentation Attention for Occluded Person Re-identification

Haofeng Meng; Qingjie Zhao

doi:10.1007/978-981-16-6372-7_74

A Lightweight Model Based on Co-segmentation Attention for Occluded Person Re-identification

Haofeng Meng, Qingjie Zhao^*

^*Corresponding author for this work

School of Computer Science and Technology

Beijing Institute of Technology

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

1 Citation (Scopus)

Abstract

Occlusion generally exists in application scenarios of person re-identification, especially in crowded situations. It is challenging to solve the occlusion problem due to the differences in the size, shape and color of the occlusion. Although pose information is helpful in solving the occlusion problem, it usually takes much time and computing resources, and its accuracy is still limited at present. To address this issue, we propose a lightweight model that uses co-segmentation attention and local features to solve the occlusion problem by attention mechanism. Experimental results on three reported datasets show that the performance of our proposed model surpasses most existing methods and is competitive with the most state-of-the-art method.

Original language	English
Title of host publication	Proceedings of 2021 Chinese Intelligent Automation Conference
Editors	Zhidong Deng
Publisher	Springer Science and Business Media Deutschland GmbH
Pages	692-701
Number of pages	10
ISBN (Print)	9789811663710
DOIs	https://doi.org/10.1007/978-981-16-6372-7_74
Publication status	Published - 2022
Event	Chinese Intelligent Automation Conference, CIAC 2021 - Zhanjiang, China Duration: 5 Nov 2021 → 7 Nov 2021

Publication series

Name	Lecture Notes in Electrical Engineering
Volume	801 LNEE
ISSN (Print)	1876-1100
ISSN (Electronic)	1876-1119

Conference

Conference	Chinese Intelligent Automation Conference, CIAC 2021
Country/Territory	China
City	Zhanjiang
Period	5/11/21 → 7/11/21

Keywords

Co-segmentation attention
Local features
Occluded person re-identification

Access to Document

10.1007/978-981-16-6372-7_74

Cite this

Meng, H., & Zhao, Q. (2022). A Lightweight Model Based on Co-segmentation Attention for Occluded Person Re-identification. In Z. Deng (Ed.), Proceedings of 2021 Chinese Intelligent Automation Conference (pp. 692-701). (Lecture Notes in Electrical Engineering; Vol. 801 LNEE). Springer Science and Business Media Deutschland GmbH. https://doi.org/10.1007/978-981-16-6372-7_74

@inproceedings{e5b8d2f0f901497bae688dd770a8aca3,

title = "A Lightweight Model Based on Co-segmentation Attention for Occluded Person Re-identification",

abstract = "Occlusion generally exists in application scenarios of person re-identification, especially in crowded situations. It is challenging to solve the occlusion problem due to the differences in the size, shape and color of the occlusion. Although pose information is helpful in solving the occlusion problem, it usually takes much time and computing resources, and its accuracy is still limited at present. To address this issue, we propose a lightweight model that uses co-segmentation attention and local features to solve the occlusion problem by attention mechanism. Experimental results on three reported datasets show that the performance of our proposed model surpasses most existing methods and is competitive with the most state-of-the-art method.",

keywords = "Co-segmentation attention, Local features, Occluded person re-identification",

author = "Haofeng Meng and Qingjie Zhao",

note = "Publisher Copyright: {\textcopyright} 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.; Chinese Intelligent Automation Conference, CIAC 2021 ; Conference date: 05-11-2021 Through 07-11-2021",

year = "2022",

doi = "10.1007/978-981-16-6372-7_74",

language = "English",

isbn = "9789811663710",

series = "Lecture Notes in Electrical Engineering",

publisher = "Springer Science and Business Media Deutschland GmbH",

pages = "692--701",

editor = "Zhidong Deng",

booktitle = "Proceedings of 2021 Chinese Intelligent Automation Conference",

address = "Germany",

}

Meng, H & Zhao, Q 2022, A Lightweight Model Based on Co-segmentation Attention for Occluded Person Re-identification. in Z Deng (ed.), Proceedings of 2021 Chinese Intelligent Automation Conference. Lecture Notes in Electrical Engineering, vol. 801 LNEE, Springer Science and Business Media Deutschland GmbH, pp. 692-701, Chinese Intelligent Automation Conference, CIAC 2021, Zhanjiang, China, 5/11/21. https://doi.org/10.1007/978-981-16-6372-7_74

A Lightweight Model Based on Co-segmentation Attention for Occluded Person Re-identification. / Meng, Haofeng; Zhao, Qingjie.
Proceedings of 2021 Chinese Intelligent Automation Conference. ed. / Zhidong Deng. Springer Science and Business Media Deutschland GmbH, 2022. p. 692-701 (Lecture Notes in Electrical Engineering; Vol. 801 LNEE).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - A Lightweight Model Based on Co-segmentation Attention for Occluded Person Re-identification

AU - Meng, Haofeng

AU - Zhao, Qingjie

PY - 2022

Y1 - 2022

N2 - Occlusion generally exists in application scenarios of person re-identification, especially in crowded situations. It is challenging to solve the occlusion problem due to the differences in the size, shape and color of the occlusion. Although pose information is helpful in solving the occlusion problem, it usually takes much time and computing resources, and its accuracy is still limited at present. To address this issue, we propose a lightweight model that uses co-segmentation attention and local features to solve the occlusion problem by attention mechanism. Experimental results on three reported datasets show that the performance of our proposed model surpasses most existing methods and is competitive with the most state-of-the-art method.

AB - Occlusion generally exists in application scenarios of person re-identification, especially in crowded situations. It is challenging to solve the occlusion problem due to the differences in the size, shape and color of the occlusion. Although pose information is helpful in solving the occlusion problem, it usually takes much time and computing resources, and its accuracy is still limited at present. To address this issue, we propose a lightweight model that uses co-segmentation attention and local features to solve the occlusion problem by attention mechanism. Experimental results on three reported datasets show that the performance of our proposed model surpasses most existing methods and is competitive with the most state-of-the-art method.

KW - Co-segmentation attention

KW - Local features

KW - Occluded person re-identification

UR - http://www.scopus.com/inward/record.url?scp=85118173331&partnerID=8YFLogxK

U2 - 10.1007/978-981-16-6372-7_74

DO - 10.1007/978-981-16-6372-7_74

M3 - Conference contribution

AN - SCOPUS:85118173331

SN - 9789811663710

T3 - Lecture Notes in Electrical Engineering

SP - 692

EP - 701

BT - Proceedings of 2021 Chinese Intelligent Automation Conference

A2 - Deng, Zhidong

PB - Springer Science and Business Media Deutschland GmbH

T2 - Chinese Intelligent Automation Conference, CIAC 2021

Y2 - 5 November 2021 through 7 November 2021

ER -