Réda Dehak

Support vector machines and joint factor analysis for speaker verification

By Najim Dehak, Patrick Kenny, Réda Dehak, Ondrej Glember, Pierre Dumouchel, Lukas Burget, Valiantsina Hubeika, Fabio Castaldo

2009-04-19

In IEEE-ICASSP

Abstract This article presents several techniques to combine between Support vector machines (SVM) and Joint Factor Analysis (JFA) model for speaker verification. In this combination, the SVMs are applied on different sources of information produced by the JFA. These informations are the Gaussian Mixture Model supervectors and speakers and Common factors. We found that the use of JFA factors gave the best results especially when within class covariance normalization method is applied in the speaker factors space, in order to compensate for the channel effect.

Continue reading

The LRDE systems for the 2008 NIST speaker recognition evaluation

By Réda Dehak, Najim Dehak, Patrick Kenny

2008-06-01

In NIST-SRE 2008

Abstract

Continue reading

Comparison between factor analysis and GMM support vector machines for speaker verification

By Najim Dehak, Réda Dehak, Patrick Kenny, Pierre Dumouchel

2007-09-25

In Proceedings of the speaker and language recognition workshop (IEEE-odyssey 2008)

Abstract We present a comparison between speaker verification systems based on factor analysis modeling and support vector machines using GMM supervectors as features. All systems used the same acoustic features and they were trained and tested on the same data sets. We test two types of kernel (one linear, the other non-linear) for the GMM support vector machines. The results show that factor analysis using speaker factors gives the best results on the core condition of the NIST 2006 speaker recognition evaluation.

Continue reading

Kernel combination for SVM speaker verification

By Réda Dehak, Najim Dehak, Patrick Kenny, Pierre Dumouchel

2007-09-25

In Proceedings of the speaker and language recognition workshop (IEEE-odyssey 2008)

Abstract We present a new approach for constructing the kernels used to build support vector machines for speaker verification. The idea is to construct new kernels by taking linear combination of many kernels such as the GLDS and GMM supervector kernels. In this new kernel combination, the combination weights are speaker dependent rather than universal weights on score level fusion and there is no need for extra-data to estimate them. An experiment on the NIST 2006 speaker recognition evaluation dataset (all trial) was done using three different kernel functions (GLDS kernel, linear and Gaussian GMM supervector kernels).

Continue reading

The role of speaker factors in the NIST extended data task

By Patrick Kenny, Najim Dehak, Réda Dehak, Vishwa Gupta, Pierre Dumouchel

2007-09-25

In Proceedings of the speaker and language recognition workshop (IEEE-odyssey 2008)

Abstract We tested factor analysis models having various numbers of speaker factors on the core condition and the extended data condition of the 2006 NIST speaker recognition evaluation. In order to ensure strict disjointness between training and test sets, the factor analysis models were trained without using any of the data made available for the 2005 evaluation. The factor analysis training set consisted primarily of Switchboard data and so was to some degree mismatched with the 2006 test data (drawn from the Mixer collection).

Continue reading

Linear and non linear kernel GMM SuperVector machines for speaker verification

By Réda Dehak, Najim Dehak, Patrick Kenny, Pierre Dumouchel

2007-08-27

In Proceedings of the european conference on speech communication and technologies (interspeech’07)

Abstract This paper presents a comparison between Support Vector Machines (SVM) speaker verification systems based on linear and non linear kernels defined in GMM supervector space. We describe how these kernel functions are related and we show how the nuisance attribute projection (NAP) technique can be used with both of these kernels to deal with the session variability problem. We demonstrate the importance of GMM model normalization (M-Norm) especially for the non linear kernel.

Continue reading

LRDE system description

By Réda Dehak, Charles-Alban Deledalle, Najim Dehak

2006-06-01

In NIST SRE’06 workshop: Speaker recognition evaluation campaign

Abstract

Continue reading

ENST-IRCGN system description

By Patrick Perrot, Réda Dehak, Gérard Chollet

2006-05-30

In NIST SRE’06 workshop: Speaker recognition evaluation campaign

Abstract

Continue reading

Spatial reasoning with relative incomplete information on relative positioning

By Réda Dehak, Isabelle Bloch, Henri Maı̂tre

2005-09-01

In IEEE Transactions on Pattern Analysis and Machine Intelligence

Abstract This paper describes a probabilistic method of inferring the position of a point with respect to a reference point knowing their relative spatial position to a third point. We address this problem in the case of incomplete information where only the angular spatial relationships are known. The use of probabilistic representations allows us to model prior knowledge. We derive exact formulae expressing the conditional probability of the position given the two known angles, in typical cases: uniform or Gaussian random prior distributions within rectangular or circular regions.

Continue reading

A novel method to fight the non-line-of-sight error in AOA measurements for mobile location

By Emmanuel Grosicki, Karim Abed-Meraim, Réda Dehak

2004-06-01

In Proceedings of the IEEE international conference on communications (ICC)

Abstract In this contribution, a mobile location method is provided using measurements from two different Base-Stations. Although computationally from two different Base-Stations. Although based on a simple trilateration and takes into account error measurements caused by Non-Line-Of-Sight (NLOS) and near-far effect. The new method attributes an index of confidence for each measure, in order to allow the mobile to select the two most reliable measures and not to use all measures, equally.

Continue reading

Search

Tags