Convolutional neural network super resolution for face recognition in surveillance monitoring

Pejman Rasti; Tõnis Uiboupin; Sergio Escalera; Gholamreza Anbarjafari

doi:10.1007/978-3-319-41778-3_18

Convolutional neural network super resolution for face recognition in surveillance monitoring

Pejman Rasti^*, Tõnis Uiboupin, Sergio Escalera, Gholamreza Anbarjafari

^*Kontaktforfatter

Publikation: Bidrag til bog/antologi/rapport/konference proceeding › Konferenceartikel i proceeding › Forskning › peer review

134 Citationer (Scopus)

Abstract

Due to the importance of security in society, monitoring activities and recognizing specific people through surveillance video cameras play an important role. One of the main issues in such activity arises from the fact that cameras do not meet the resolution requirement for many face recognition algorithms. In order to solve this issue, in this paper we are proposing a new system which super resolves the image using deep learning convolutional network followed by the Hidden Markov Model and Singular Value Decomposition based face recognition. The proposed system has been tested on many well-known face databases such as FERET, HeadPose, and Essex University databases as well as our recently introduced iCV Face Recognition database (iCVF). The experimental results show that the recognition rate is improving considerably after apply the super resolution.

Originalsprog	Engelsk
Titel	Articulated Motion and Deformable Objects - 9th International Conference, AMDO 2016, Proceedings
Redaktører	Francisco José Perales, Josef Kittler
Antal sider	10
Forlag	Physica-Verlag
Publikationsdato	2016
Sider	175-184
ISBN (Trykt)	9783319417776
DOI	https://doi.org/10.1007/978-3-319-41778-3_18
Status	Udgivet - 2016
Udgivet eksternt	Ja
Begivenhed	9th International Conference on Articulated Motion and Deformable Objects, AMDO 2016 - Palma de Mallorca, Spanien Varighed: 13 jul. 2016 → 15 jul. 2016

Konference

Konference	9th International Conference on Articulated Motion and Deformable Objects, AMDO 2016
Land/Område	Spanien
By	Palma de Mallorca
Periode	13/07/2016 → 15/07/2016
Sponsor	AERFAI (Spanish Association for Pattern Recognition and Image Analysis), Ajuntament de Palma, BMN, Banco Mare Nostrum, S.A, Escola Politecnica Superior (UIB) - Universitat de les Illes Balears (UIB), Mathematics and Computer Science Department, Universitat de les Illes Balears (UIB)

Navn	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Vol/bind	9756
ISSN	0302-9743

Bibliografisk note

Publisher Copyright:
© Springer International Publishing Switzerland 2016.

Adgang til dokumentet

10.1007/978-3-319-41778-3_18

AUB Link

Søg efter materialet i Aalborg Universitetsbiblioteks søgemaskine

Andre filer og links

Link to publication in Scopus

Citationsformater

Rasti, Pejman ; Uiboupin, Tõnis ; Escalera, Sergio et al. / Convolutional neural network super resolution for face recognition in surveillance monitoring. Articulated Motion and Deformable Objects - 9th International Conference, AMDO 2016, Proceedings. red. / Francisco José Perales ; Josef Kittler. Physica-Verlag, 2016. s. 175-184 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bind 9756).

@inproceedings{4d3feda8b96f4b3f92d836154aca8aeb,

title = "Convolutional neural network super resolution for face recognition in surveillance monitoring",

abstract = "Due to the importance of security in society, monitoring activities and recognizing specific people through surveillance video cameras play an important role. One of the main issues in such activity arises from the fact that cameras do not meet the resolution requirement for many face recognition algorithms. In order to solve this issue, in this paper we are proposing a new system which super resolves the image using deep learning convolutional network followed by the Hidden Markov Model and Singular Value Decomposition based face recognition. The proposed system has been tested on many well-known face databases such as FERET, HeadPose, and Essex University databases as well as our recently introduced iCV Face Recognition database (iCVF). The experimental results show that the recognition rate is improving considerably after apply the super resolution.",

keywords = "Deep learning, Face recognition, Hidden markov model, Super resolution, Support vector machine, Surveillance videos",

author = "Pejman Rasti and T{\~o}nis Uiboupin and Sergio Escalera and Gholamreza Anbarjafari",

note = "Publisher Copyright: {\textcopyright} Springer International Publishing Switzerland 2016.; 9th International Conference on Articulated Motion and Deformable Objects, AMDO 2016 ; Conference date: 13-07-2016 Through 15-07-2016",

year = "2016",

doi = "10.1007/978-3-319-41778-3_18",

language = "English",

isbn = "9783319417776",

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

publisher = "Physica-Verlag",

pages = "175--184",

editor = "Perales, {Francisco Jos{\'e}} and Josef Kittler",

booktitle = "Articulated Motion and Deformable Objects - 9th International Conference, AMDO 2016, Proceedings",

}

Rasti, P, Uiboupin, T, Escalera, S & Anbarjafari, G 2016, Convolutional neural network super resolution for face recognition in surveillance monitoring. i FJ Perales & J Kittler (red), Articulated Motion and Deformable Objects - 9th International Conference, AMDO 2016, Proceedings. Physica-Verlag, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), bind 9756, s. 175-184, 9th International Conference on Articulated Motion and Deformable Objects, AMDO 2016, Palma de Mallorca, Spanien, 13/07/2016. https://doi.org/10.1007/978-3-319-41778-3_18

Convolutional neural network super resolution for face recognition in surveillance monitoring. / Rasti, Pejman; Uiboupin, Tõnis; Escalera, Sergio et al.
Articulated Motion and Deformable Objects - 9th International Conference, AMDO 2016, Proceedings. red. / Francisco José Perales; Josef Kittler. Physica-Verlag, 2016. s. 175-184 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bind 9756).

Publikation: Bidrag til bog/antologi/rapport/konference proceeding › Konferenceartikel i proceeding › Forskning › peer review

TY - GEN

T1 - Convolutional neural network super resolution for face recognition in surveillance monitoring

AU - Rasti, Pejman

AU - Uiboupin, Tõnis

AU - Escalera, Sergio

AU - Anbarjafari, Gholamreza

N1 - Publisher Copyright: © Springer International Publishing Switzerland 2016.

PY - 2016

Y1 - 2016

N2 - Due to the importance of security in society, monitoring activities and recognizing specific people through surveillance video cameras play an important role. One of the main issues in such activity arises from the fact that cameras do not meet the resolution requirement for many face recognition algorithms. In order to solve this issue, in this paper we are proposing a new system which super resolves the image using deep learning convolutional network followed by the Hidden Markov Model and Singular Value Decomposition based face recognition. The proposed system has been tested on many well-known face databases such as FERET, HeadPose, and Essex University databases as well as our recently introduced iCV Face Recognition database (iCVF). The experimental results show that the recognition rate is improving considerably after apply the super resolution.

AB - Due to the importance of security in society, monitoring activities and recognizing specific people through surveillance video cameras play an important role. One of the main issues in such activity arises from the fact that cameras do not meet the resolution requirement for many face recognition algorithms. In order to solve this issue, in this paper we are proposing a new system which super resolves the image using deep learning convolutional network followed by the Hidden Markov Model and Singular Value Decomposition based face recognition. The proposed system has been tested on many well-known face databases such as FERET, HeadPose, and Essex University databases as well as our recently introduced iCV Face Recognition database (iCVF). The experimental results show that the recognition rate is improving considerably after apply the super resolution.

KW - Deep learning

KW - Face recognition

KW - Hidden markov model

KW - Super resolution

KW - Support vector machine

KW - Surveillance videos

UR - http://www.scopus.com/inward/record.url?scp=84978204107&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-41778-3_18

DO - 10.1007/978-3-319-41778-3_18

M3 - Article in proceeding

AN - SCOPUS:84978204107

SN - 9783319417776

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

SP - 175

EP - 184

BT - Articulated Motion and Deformable Objects - 9th International Conference, AMDO 2016, Proceedings

A2 - Perales, Francisco José

A2 - Kittler, Josef

PB - Physica-Verlag

T2 - 9th International Conference on Articulated Motion and Deformable Objects, AMDO 2016

Y2 - 13 July 2016 through 15 July 2016

ER -

Rasti P, Uiboupin T, Escalera S, Anbarjafari G. Convolutional neural network super resolution for face recognition in surveillance monitoring. I Perales FJ, Kittler J, red., Articulated Motion and Deformable Objects - 9th International Conference, AMDO 2016, Proceedings. Physica-Verlag. 2016. s. 175-184. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Bind 9756). doi: 10.1007/978-3-319-41778-3_18

Convolutional neural network super resolution for face recognition in surveillance monitoring

Abstract

Konference

Bibliografisk note

Adgang til dokumentet

AUB Link

Andre filer og links

Fingeraftryk

Citationsformater