A deep neural network for audio-visual person recognition

M.R. Alam, Mohammed Bennamoun, Roberto Togneri, Ferdous Sohel

Research output: Chapter in Book/Conference paper › Conference paper

6 Citations (Scopus)

Abstract

© 2015 IEEE. This paper presents applications of special types of deep neural networks (DNNs) for audio-visual biometrics. A common example is the DBN-DNN that uses the generative weights of deep belief networks (DBNs) to initialize the feature detecting layers of deterministic feed forward DNNs. In this paper, we propose the DBM-DNN that uses the generative weights of deep Boltzmann machines (DBMs) for initialization of DNNs. Then, a softmax layer is added on top and the DNNs are trained discriminatively. Our experimental results show that lower error rates can be achieved using the DBM-DNN compared to the support vector machine (SVM), linear regression-based classifier (LRC) and the DBN-DNN. Experiments were carried out on two publicly available audio-visual datasets: the VidTIMIT and MOBIO.
Original languageEnglish
Title of host publicationBiometrics Theory, Applications and Systems (BTAS), 2015 IEEE 7th International Conference
Place of PublicationUSA
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages1-6
VolumeN/A
ISBN (Print)9781479987764
DOIs
Publication statusPublished - 2015
EventBiometrics Theory, Applications and Systems (BTAS) 2015 - Virginia, United States
Duration: 8 Sep 2015 → 11 Sep 2015

Conference

ConferenceBiometrics Theory, Applications and Systems (BTAS) 2015
CountryUnited States
CityVirginia
Period8/09/15 → 11/09/15

Fingerprint

Bayesian networks
Deep neural networks
Biometrics
Linear regression
Support vector machines
Classifiers
Experiments

Cite this

Alam, M. R., Bennamoun, M., Togneri, R., & Sohel, F. (2015). A deep neural network for audio-visual person recognition. In Biometrics Theory, Applications and Systems (BTAS), 2015 IEEE 7th International Conference (Vol. N/A, pp. 1-6). USA: IEEE, Institute of Electrical and Electronics Engineers. https://doi.org/10.1109/BTAS.2015.7358754
Alam, M.R. ; Bennamoun, Mohammed ; Togneri, Roberto ; Sohel, Ferdous. / A deep neural network for audio-visual person recognition. Biometrics Theory, Applications and Systems (BTAS), 2015 IEEE 7th International Conference . Vol. N/A USA : IEEE, Institute of Electrical and Electronics Engineers, 2015. pp. 1-6
@inproceedings{d10bc81c82bf45ccb3c96b45a1457a73,
title = "A deep neural network for audio-visual person recognition",
abstract = "{\circledC} 2015 IEEE. This paper presents applications of special types of deep neural networks (DNNs) for audio-visual biometrics. A common example is the DBN-DNN that uses the generative weights of deep belief networks (DBNs) to initialize the feature detecting layers of deterministic feed forward DNNs. In this paper, we propose the DBM-DNN that uses the generative weights of deep Boltzmann machines (DBMs) for initialization of DNNs. Then, a softmax layer is added on top and the DNNs are trained discriminatively. Our experimental results show that lower error rates can be achieved using the DBM-DNN compared to the support vector machine (SVM), linear regression-based classifier (LRC) and the DBN-DNN. Experiments were carried out on two publicly available audio-visual datasets: the VidTIMIT and MOBIO.",
author = "M.R. Alam and Mohammed Bennamoun and Roberto Togneri and Ferdous Sohel",
year = "2015",
doi = "10.1109/BTAS.2015.7358754",
language = "English",
isbn = "9781479987764",
volume = "N/A",
pages = "1--6",
booktitle = "Biometrics Theory, Applications and Systems (BTAS), 2015 IEEE 7th International Conference",
publisher = "IEEE, Institute of Electrical and Electronics Engineers",
address = "United States",

}

Alam, MR, Bennamoun, M, Togneri, R & Sohel, F 2015, A deep neural network for audio-visual person recognition. in Biometrics Theory, Applications and Systems (BTAS), 2015 IEEE 7th International Conference . vol. N/A, IEEE, Institute of Electrical and Electronics Engineers, USA, pp. 1-6, Biometrics Theory, Applications and Systems (BTAS) 2015, Virginia, United States, 8/09/15. https://doi.org/10.1109/BTAS.2015.7358754

A deep neural network for audio-visual person recognition. / Alam, M.R.; Bennamoun, Mohammed; Togneri, Roberto; Sohel, Ferdous.

Biometrics Theory, Applications and Systems (BTAS), 2015 IEEE 7th International Conference . Vol. N/A USA : IEEE, Institute of Electrical and Electronics Engineers, 2015. p. 1-6.

Research output: Chapter in Book/Conference paper › Conference paper

TY - GEN

T1 - A deep neural network for audio-visual person recognition

AU - Alam, M.R.

AU - Bennamoun, Mohammed

AU - Togneri, Roberto

AU - Sohel, Ferdous

PY - 2015

Y1 - 2015

N2 - © 2015 IEEE. This paper presents applications of special types of deep neural networks (DNNs) for audio-visual biometrics. A common example is the DBN-DNN that uses the generative weights of deep belief networks (DBNs) to initialize the feature detecting layers of deterministic feed forward DNNs. In this paper, we propose the DBM-DNN that uses the generative weights of deep Boltzmann machines (DBMs) for initialization of DNNs. Then, a softmax layer is added on top and the DNNs are trained discriminatively. Our experimental results show that lower error rates can be achieved using the DBM-DNN compared to the support vector machine (SVM), linear regression-based classifier (LRC) and the DBN-DNN. Experiments were carried out on two publicly available audio-visual datasets: the VidTIMIT and MOBIO.

AB - © 2015 IEEE. This paper presents applications of special types of deep neural networks (DNNs) for audio-visual biometrics. A common example is the DBN-DNN that uses the generative weights of deep belief networks (DBNs) to initialize the feature detecting layers of deterministic feed forward DNNs. In this paper, we propose the DBM-DNN that uses the generative weights of deep Boltzmann machines (DBMs) for initialization of DNNs. Then, a softmax layer is added on top and the DNNs are trained discriminatively. Our experimental results show that lower error rates can be achieved using the DBM-DNN compared to the support vector machine (SVM), linear regression-based classifier (LRC) and the DBN-DNN. Experiments were carried out on two publicly available audio-visual datasets: the VidTIMIT and MOBIO.

U2 - 10.1109/BTAS.2015.7358754

DO - 10.1109/BTAS.2015.7358754

M3 - Conference paper

SN - 9781479987764

VL - N/A

SP - 1

EP - 6

BT - Biometrics Theory, Applications and Systems (BTAS), 2015 IEEE 7th International Conference

PB - IEEE, Institute of Electrical and Electronics Engineers

CY - USA

ER -

Alam MR, Bennamoun M, Togneri R, Sohel F. A deep neural network for audio-visual person recognition. In Biometrics Theory, Applications and Systems (BTAS), 2015 IEEE 7th International Conference . Vol. N/A. USA: IEEE, Institute of Electrical and Electronics Engineers. 2015. p. 1-6 https://doi.org/10.1109/BTAS.2015.7358754

Access to Document