Improved RGB-D-T based Face Recognition

Marc Oliu Simon, Ciprian Corneanu, Kamal Nasrollahi, Olegs Nikisins, Sergio Escalera Guerrero, Yunlian Sun, Haiqing Li, Zhenan Sun, Thomas B. Moeslund, Modris Greitans

Research output: Contribution to journalJournal articleResearchpeer-review

18 Citations (Scopus)
827 Downloads (Pure)

Abstract

Reliable facial recognition systems are of crucial importance in various applications from entertainment to security. Thanks to the deep-learning concepts introduced in the field, a significant improvement in the performance of the unimodal facial recognition systems has been observed in the recent years. At the same time a multimodal facial recognition is a promising approach. This paper combines the latest successes in both directions by applying deep learning Convolutional Neural Networks (CNN) to the multimodal RGB-D-T based facial recognition problem outperforming previously published results. Furthermore, a late fusion of the CNN-based recognition block with various hand-crafted features (LBP, HOG, HAAR, HOGOM) is introduced, demonstrating even better recognition performance on a benchmark RGB-D-T database. The obtained results in this paper show that the classical engineered features and CNN-based features can complement each other for recognition purposes.
Original languageEnglish
JournalIET Biometrics
Volume5
Issue number4
Pages (from-to)297 - 303
ISSN2047-4946
DOIs
Publication statusPublished - 2016

Fingerprint

Face recognition
Neural networks
Fusion reactions
Deep learning

Keywords

  • multimodal
  • face recognition
  • RGB
  • depth
  • thermal

Cite this

Oliu Simon, M., Corneanu, C., Nasrollahi, K., Nikisins, O., Guerrero, S. E., Sun, Y., ... Greitans, M. (2016). Improved RGB-D-T based Face Recognition. IET Biometrics, 5(4), 297 - 303. https://doi.org/10.1049/iet-bmt.2015.0057
Oliu Simon, Marc ; Corneanu, Ciprian ; Nasrollahi, Kamal ; Nikisins, Olegs ; Guerrero, Sergio Escalera ; Sun, Yunlian ; Li, Haiqing ; Sun, Zhenan ; Moeslund, Thomas B. ; Greitans, Modris. / Improved RGB-D-T based Face Recognition. In: IET Biometrics. 2016 ; Vol. 5, No. 4. pp. 297 - 303.
@article{6ff0ce1896b846f386dab78519738545,
title = "Improved RGB-D-T based Face Recognition",
abstract = "Reliable facial recognition systems are of crucial importance in various applications from entertainment to security. Thanks to the deep-learning concepts introduced in the field, a significant improvement in the performance of the unimodal facial recognition systems has been observed in the recent years. At the same time a multimodal facial recognition is a promising approach. This paper combines the latest successes in both directions by applying deep learning Convolutional Neural Networks (CNN) to the multimodal RGB-D-T based facial recognition problem outperforming previously published results. Furthermore, a late fusion of the CNN-based recognition block with various hand-crafted features (LBP, HOG, HAAR, HOGOM) is introduced, demonstrating even better recognition performance on a benchmark RGB-D-T database. The obtained results in this paper show that the classical engineered features and CNN-based features can complement each other for recognition purposes.",
keywords = "multimodal, face recognition, RGB, depth, thermal",
author = "{Oliu Simon}, Marc and Ciprian Corneanu and Kamal Nasrollahi and Olegs Nikisins and Guerrero, {Sergio Escalera} and Yunlian Sun and Haiqing Li and Zhenan Sun and Moeslund, {Thomas B.} and Modris Greitans",
year = "2016",
doi = "10.1049/iet-bmt.2015.0057",
language = "English",
volume = "5",
pages = "297 -- 303",
journal = "IET Biometrics",
issn = "2047-4938",
publisher = "TheInstitution of Engineering and Technology",
number = "4",

}

Oliu Simon, M, Corneanu, C, Nasrollahi, K, Nikisins, O, Guerrero, SE, Sun, Y, Li, H, Sun, Z, Moeslund, TB & Greitans, M 2016, 'Improved RGB-D-T based Face Recognition', IET Biometrics, vol. 5, no. 4, pp. 297 - 303. https://doi.org/10.1049/iet-bmt.2015.0057

Improved RGB-D-T based Face Recognition. / Oliu Simon, Marc; Corneanu, Ciprian; Nasrollahi, Kamal; Nikisins, Olegs; Guerrero, Sergio Escalera; Sun, Yunlian; Li, Haiqing; Sun, Zhenan; Moeslund, Thomas B.; Greitans, Modris.

In: IET Biometrics, Vol. 5, No. 4, 2016, p. 297 - 303.

Research output: Contribution to journalJournal articleResearchpeer-review

TY - JOUR

T1 - Improved RGB-D-T based Face Recognition

AU - Oliu Simon, Marc

AU - Corneanu, Ciprian

AU - Nasrollahi, Kamal

AU - Nikisins, Olegs

AU - Guerrero, Sergio Escalera

AU - Sun, Yunlian

AU - Li, Haiqing

AU - Sun, Zhenan

AU - Moeslund, Thomas B.

AU - Greitans, Modris

PY - 2016

Y1 - 2016

N2 - Reliable facial recognition systems are of crucial importance in various applications from entertainment to security. Thanks to the deep-learning concepts introduced in the field, a significant improvement in the performance of the unimodal facial recognition systems has been observed in the recent years. At the same time a multimodal facial recognition is a promising approach. This paper combines the latest successes in both directions by applying deep learning Convolutional Neural Networks (CNN) to the multimodal RGB-D-T based facial recognition problem outperforming previously published results. Furthermore, a late fusion of the CNN-based recognition block with various hand-crafted features (LBP, HOG, HAAR, HOGOM) is introduced, demonstrating even better recognition performance on a benchmark RGB-D-T database. The obtained results in this paper show that the classical engineered features and CNN-based features can complement each other for recognition purposes.

AB - Reliable facial recognition systems are of crucial importance in various applications from entertainment to security. Thanks to the deep-learning concepts introduced in the field, a significant improvement in the performance of the unimodal facial recognition systems has been observed in the recent years. At the same time a multimodal facial recognition is a promising approach. This paper combines the latest successes in both directions by applying deep learning Convolutional Neural Networks (CNN) to the multimodal RGB-D-T based facial recognition problem outperforming previously published results. Furthermore, a late fusion of the CNN-based recognition block with various hand-crafted features (LBP, HOG, HAAR, HOGOM) is introduced, demonstrating even better recognition performance on a benchmark RGB-D-T database. The obtained results in this paper show that the classical engineered features and CNN-based features can complement each other for recognition purposes.

KW - multimodal

KW - face recognition

KW - RGB

KW - depth

KW - thermal

U2 - 10.1049/iet-bmt.2015.0057

DO - 10.1049/iet-bmt.2015.0057

M3 - Journal article

VL - 5

SP - 297

EP - 303

JO - IET Biometrics

JF - IET Biometrics

SN - 2047-4938

IS - 4

ER -