Similarity Scores Based Re-classification for Open-Set Person Re-identification

Hongsheng Wang; Yang Yang; Shengcai Liao; Dong Cao; Zhen Lei

doi:10.1007/978-3-030-31456-9_54

Similarity Scores Based Re-classification for Open-Set Person Re-identification

Hongsheng Wang, Yang Yang, Shengcai Liao, Dong Cao, Zhen Lei

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

Abstract

In this paper, we propose a new similarity scores based re-classification method for open-set person re-identification, which exploits information among the top-n most similar matching candidates in the gallery set. Moreover, to make the cross-view quadratic discriminant analysis metric learning method effectively learn both the projection matrix and the metric kernel with open-set data, we introduce an additional regularization factor to adjust the covariance matrix of the obtained subspace. Our Experiments on challenging OPeRID v1.0 database show that our approach improves the Rank-1 recognition rates at 1% FAR by 8.86% and 10.51% with re-ranking, respectively.

Original language	English
Title of host publication	Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings
Editors	Zhenan Sun, Ran He, Shiguang Shan, Jianjiang Feng, Zhenhua Guo
Publisher	Springer
Pages	493-501
Number of pages	9
ISBN (Print)	9783030314552
DOIs	https://doi.org/10.1007/978-3-030-31456-9_54
Publication status	Published - 2019
Externally published	Yes
Event	14th Chinese Conference on Biometric Recognition, CCBR 2019 - Zhuzhou, China Duration: Oct 12 2019 → Oct 13 2019

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	11818 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	14th Chinese Conference on Biometric Recognition, CCBR 2019
Country/Territory	China
City	Zhuzhou
Period	10/12/19 → 10/13/19

Keywords

Metric learning
Open-set
Person re-identification
Re-classification

ASJC Scopus subject areas

Theoretical Computer Science
General Computer Science

Access to Document

10.1007/978-3-030-31456-9_54

Cite this

Wang, H., Yang, Y., Liao, S., Cao, D., & Lei, Z. (2019). Similarity Scores Based Re-classification for Open-Set Person Re-identification. In Z. Sun, R. He, S. Shan, J. Feng, & Z. Guo (Eds.), Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings (pp. 493-501). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11818 LNCS). Springer. https://doi.org/10.1007/978-3-030-31456-9_54

Similarity Scores Based Re-classification for Open-Set Person Re-identification. / Wang, Hongsheng; Yang, Yang; Liao, Shengcai et al.
Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. ed. / Zhenan Sun; Ran He; Shiguang Shan; Jianjiang Feng; Zhenhua Guo. Springer, 2019. p. 493-501 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 11818 LNCS).

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

Wang, H, Yang, Y, Liao, S, Cao, D & Lei, Z 2019, Similarity Scores Based Re-classification for Open-Set Person Re-identification. in Z Sun, R He, S Shan, J Feng & Z Guo (eds), Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 11818 LNCS, Springer, pp. 493-501, 14th Chinese Conference on Biometric Recognition, CCBR 2019, Zhuzhou, China, 10/12/19. https://doi.org/10.1007/978-3-030-31456-9_54

Wang H, Yang Y, Liao S, Cao D, Lei Z. Similarity Scores Based Re-classification for Open-Set Person Re-identification. In Sun Z, He R, Shan S, Feng J, Guo Z, editors, Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. Springer. 2019. p. 493-501. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-030-31456-9_54

Wang, Hongsheng ; Yang, Yang ; Liao, Shengcai et al. / Similarity Scores Based Re-classification for Open-Set Person Re-identification. Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings. editor / Zhenan Sun ; Ran He ; Shiguang Shan ; Jianjiang Feng ; Zhenhua Guo. Springer, 2019. pp. 493-501 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{3371f9240f9b48f887784554cd5f28a9,

title = "Similarity Scores Based Re-classification for Open-Set Person Re-identification",

abstract = "In this paper, we propose a new similarity scores based re-classification method for open-set person re-identification, which exploits information among the top-n most similar matching candidates in the gallery set. Moreover, to make the cross-view quadratic discriminant analysis metric learning method effectively learn both the projection matrix and the metric kernel with open-set data, we introduce an additional regularization factor to adjust the covariance matrix of the obtained subspace. Our Experiments on challenging OPeRID v1.0 database show that our approach improves the Rank-1 recognition rates at 1% FAR by 8.86% and 10.51% with re-ranking, respectively.",

keywords = "Metric learning, Open-set, Person re-identification, Re-classification",

author = "Hongsheng Wang and Yang Yang and Shengcai Liao and Dong Cao and Zhen Lei",

note = "Publisher Copyright: {\textcopyright} 2019, Springer Nature Switzerland AG.; 14th Chinese Conference on Biometric Recognition, CCBR 2019 ; Conference date: 12-10-2019 Through 13-10-2019",

year = "2019",

doi = "10.1007/978-3-030-31456-9_54",

language = "English",

isbn = "9783030314552",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer",

pages = "493--501",

editor = "Zhenan Sun and Ran He and Shiguang Shan and Jianjiang Feng and Zhenhua Guo",

booktitle = "Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings",

}

TY - GEN

T1 - Similarity Scores Based Re-classification for Open-Set Person Re-identification

AU - Wang, Hongsheng

AU - Yang, Yang

AU - Liao, Shengcai

AU - Cao, Dong

AU - Lei, Zhen

PY - 2019

Y1 - 2019

N2 - In this paper, we propose a new similarity scores based re-classification method for open-set person re-identification, which exploits information among the top-n most similar matching candidates in the gallery set. Moreover, to make the cross-view quadratic discriminant analysis metric learning method effectively learn both the projection matrix and the metric kernel with open-set data, we introduce an additional regularization factor to adjust the covariance matrix of the obtained subspace. Our Experiments on challenging OPeRID v1.0 database show that our approach improves the Rank-1 recognition rates at 1% FAR by 8.86% and 10.51% with re-ranking, respectively.

AB - In this paper, we propose a new similarity scores based re-classification method for open-set person re-identification, which exploits information among the top-n most similar matching candidates in the gallery set. Moreover, to make the cross-view quadratic discriminant analysis metric learning method effectively learn both the projection matrix and the metric kernel with open-set data, we introduce an additional regularization factor to adjust the covariance matrix of the obtained subspace. Our Experiments on challenging OPeRID v1.0 database show that our approach improves the Rank-1 recognition rates at 1% FAR by 8.86% and 10.51% with re-ranking, respectively.

KW - Metric learning

KW - Open-set

KW - Person re-identification

KW - Re-classification

UR - http://www.scopus.com/inward/record.url?scp=85075583897&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85075583897&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-31456-9_54

DO - 10.1007/978-3-030-31456-9_54

M3 - Conference contribution

AN - SCOPUS:85075583897

SN - 9783030314552

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 493

EP - 501

BT - Biometric Recognition - 14th Chinese Conference, CCBR 2019, Proceedings

A2 - Sun, Zhenan

A2 - He, Ran

A2 - Shan, Shiguang

A2 - Feng, Jianjiang

A2 - Guo, Zhenhua

PB - Springer

T2 - 14th Chinese Conference on Biometric Recognition, CCBR 2019

Y2 - 12 October 2019 through 13 October 2019

ER -

Similarity Scores Based Re-classification for Open-Set Person Re-identification

Abstract

Publication series

Conference

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this