Vehicle verification based on deep siamese network with similarity metric

Qian Zhang; Mingtao Pei; Mei Chen; Yunde Jia

doi:10.1007/978-3-319-77380-3_74

Vehicle verification based on deep siamese network with similarity metric

Qian Zhang, Mingtao Pei^*, Mei Chen, Yunde Jia

^*Corresponding author for this work

School of Computer Science and Technology

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

4 Citations (Scopus)

Abstract

Vehicle verification is a challenging research problem with important practical applications. Most prior work focused on either feature learning or distance metric learning, which could not guarantee the compatibility of the learned feature and the distance metric. In this paper, we propose an end-to-end model based on the Siamese Convolutional Neural Network (CNN), which integrates distance metric learning and feature learning into a unified framework. The network is trained by contrastive loss and a similarity metric loss defined by joint Bayesian to learn more discriminative features for vehicle verification. The experimental results demonstrate the effectiveness of the proposed method.

Original language	English
Title of host publication	Advances in Multimedia Information Processing – PCM 2017 - 18th Pacific-Rim Conference on Multimedia, Revised Selected Papers
Editors	Bing Zeng, Hongliang Li, Abdulmotaleb El Saddik, Xiaopeng Fan, Shuqiang Jiang, Qingming Huang
Publisher	Springer Verlag
Pages	773-782
Number of pages	10
ISBN (Print)	9783319773797
DOIs	https://doi.org/10.1007/978-3-319-77380-3_74
Publication status	Published - 2018
Event	18th Pacific-Rim Conference on Multimedia, PCM 2017 - Harbin, China Duration: 28 Sept 2017 → 29 Sept 2017

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	10735 LNCS
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	18th Pacific-Rim Conference on Multimedia, PCM 2017
Country/Territory	China
City	Harbin
Period	28/09/17 → 29/09/17

Keywords

Joint bayesian
Siamese network
Similarity metric
Vehicle verification

Access to Document

10.1007/978-3-319-77380-3_74

Cite this

Zhang, Q., Pei, M., Chen, M., & Jia, Y. (2018). Vehicle verification based on deep siamese network with similarity metric. In B. Zeng, H. Li, A. El Saddik, X. Fan, S. Jiang, & Q. Huang (Eds.), Advances in Multimedia Information Processing – PCM 2017 - 18th Pacific-Rim Conference on Multimedia, Revised Selected Papers (pp. 773-782). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10735 LNCS). Springer Verlag. https://doi.org/10.1007/978-3-319-77380-3_74

Zhang, Qian ; Pei, Mingtao ; Chen, Mei et al. / Vehicle verification based on deep siamese network with similarity metric. Advances in Multimedia Information Processing – PCM 2017 - 18th Pacific-Rim Conference on Multimedia, Revised Selected Papers. editor / Bing Zeng ; Hongliang Li ; Abdulmotaleb El Saddik ; Xiaopeng Fan ; Shuqiang Jiang ; Qingming Huang. Springer Verlag, 2018. pp. 773-782 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{e95b4c657c97429ca839874edf83214c,

title = "Vehicle verification based on deep siamese network with similarity metric",

abstract = "Vehicle verification is a challenging research problem with important practical applications. Most prior work focused on either feature learning or distance metric learning, which could not guarantee the compatibility of the learned feature and the distance metric. In this paper, we propose an end-to-end model based on the Siamese Convolutional Neural Network (CNN), which integrates distance metric learning and feature learning into a unified framework. The network is trained by contrastive loss and a similarity metric loss defined by joint Bayesian to learn more discriminative features for vehicle verification. The experimental results demonstrate the effectiveness of the proposed method.",

keywords = "Joint bayesian, Siamese network, Similarity metric, Vehicle verification",

author = "Qian Zhang and Mingtao Pei and Mei Chen and Yunde Jia",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing AG, part of Springer Nature 2018.; 18th Pacific-Rim Conference on Multimedia, PCM 2017 ; Conference date: 28-09-2017 Through 29-09-2017",

year = "2018",

doi = "10.1007/978-3-319-77380-3\_74",

language = "English",

isbn = "9783319773797",

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

publisher = "Springer Verlag",

pages = "773--782",

editor = "Bing Zeng and Hongliang Li and \{El Saddik\}, Abdulmotaleb and Xiaopeng Fan and Shuqiang Jiang and Qingming Huang",

booktitle = "Advances in Multimedia Information Processing – PCM 2017 - 18th Pacific-Rim Conference on Multimedia, Revised Selected Papers",

address = "Germany",

}

Zhang, Q, Pei, M, Chen, M & Jia, Y 2018, Vehicle verification based on deep siamese network with similarity metric. in B Zeng, H Li, A El Saddik, X Fan, S Jiang & Q Huang (eds), Advances in Multimedia Information Processing – PCM 2017 - 18th Pacific-Rim Conference on Multimedia, Revised Selected Papers. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 10735 LNCS, Springer Verlag, pp. 773-782, 18th Pacific-Rim Conference on Multimedia, PCM 2017, Harbin, China, 28/09/17. https://doi.org/10.1007/978-3-319-77380-3_74

Vehicle verification based on deep siamese network with similarity metric. / Zhang, Qian; Pei, Mingtao; Chen, Mei et al.
Advances in Multimedia Information Processing – PCM 2017 - 18th Pacific-Rim Conference on Multimedia, Revised Selected Papers. ed. / Bing Zeng; Hongliang Li; Abdulmotaleb El Saddik; Xiaopeng Fan; Shuqiang Jiang; Qingming Huang. Springer Verlag, 2018. p. 773-782 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 10735 LNCS).

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

TY - GEN

T1 - Vehicle verification based on deep siamese network with similarity metric

AU - Zhang, Qian

AU - Pei, Mingtao

AU - Chen, Mei

AU - Jia, Yunde

N1 - Publisher Copyright: © Springer International Publishing AG, part of Springer Nature 2018.

PY - 2018

Y1 - 2018

N2 - Vehicle verification is a challenging research problem with important practical applications. Most prior work focused on either feature learning or distance metric learning, which could not guarantee the compatibility of the learned feature and the distance metric. In this paper, we propose an end-to-end model based on the Siamese Convolutional Neural Network (CNN), which integrates distance metric learning and feature learning into a unified framework. The network is trained by contrastive loss and a similarity metric loss defined by joint Bayesian to learn more discriminative features for vehicle verification. The experimental results demonstrate the effectiveness of the proposed method.

AB - Vehicle verification is a challenging research problem with important practical applications. Most prior work focused on either feature learning or distance metric learning, which could not guarantee the compatibility of the learned feature and the distance metric. In this paper, we propose an end-to-end model based on the Siamese Convolutional Neural Network (CNN), which integrates distance metric learning and feature learning into a unified framework. The network is trained by contrastive loss and a similarity metric loss defined by joint Bayesian to learn more discriminative features for vehicle verification. The experimental results demonstrate the effectiveness of the proposed method.

KW - Joint bayesian

KW - Siamese network

KW - Similarity metric

KW - Vehicle verification

UR - https://www.scopus.com/pages/publications/85047446334

U2 - 10.1007/978-3-319-77380-3_74

DO - 10.1007/978-3-319-77380-3_74

M3 - Conference contribution

AN - SCOPUS:85047446334

SN - 9783319773797

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

SP - 773

EP - 782

BT - Advances in Multimedia Information Processing – PCM 2017 - 18th Pacific-Rim Conference on Multimedia, Revised Selected Papers

A2 - Zeng, Bing

A2 - Li, Hongliang

A2 - El Saddik, Abdulmotaleb

A2 - Fan, Xiaopeng

A2 - Jiang, Shuqiang

A2 - Huang, Qingming

PB - Springer Verlag

T2 - 18th Pacific-Rim Conference on Multimedia, PCM 2017

Y2 - 28 September 2017 through 29 September 2017

ER -

Zhang Q, Pei M, Chen M, Jia Y. Vehicle verification based on deep siamese network with similarity metric. In Zeng B, Li H, El Saddik A, Fan X, Jiang S, Huang Q, editors, Advances in Multimedia Information Processing – PCM 2017 - 18th Pacific-Rim Conference on Multimedia, Revised Selected Papers. Springer Verlag. 2018. p. 773-782. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-77380-3_74

Vehicle verification based on deep siamese network with similarity metric

Abstract

Publication series

Conference

Keywords

Access to Document

Other files and links

Fingerprint

Cite this