Linear and Non Linear Kernel GMM SuperVector Machines for Speaker Verification

Authors

Réda Dehak, Najim Dehak, Patrick Kenny, Pierre Dumouchel

Where

Proceedings of the European Conference on Speech Communication and Technologies (Interspeech'07)

Place

Antwerp, Belgium

Type

inproceedings

Date

2007-08-27

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. All our experiments were performed on the core condition of NIST 2006 speaker recognition evaluation (all trials). Our best results (an equal error rate of 6.3%) were obtained using NAP and GMM model normalization with the non linear kernel.

Bibtex (lrde.bib)

@InProceedings{	  dehak.07.interspeech,
  author	= {R\'eda Dehak and Najim Dehak and Patrick Kenny and Pierre
		  Dumouchel},
  title		= {Linear and Non Linear Kernel {GMM} SuperVector Machines
		  for Speaker Verification},
  booktitle	= {Proceedings of the European Conference on Speech
		  Communication and Technologies (Interspeech'07)},
  year		= 2007,
  address	= {Antwerp, Belgium},
  month		= aug,
  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. All our experiments were performed on the core
		  condition of NIST 2006 speaker recognition evaluation (all
		  trials). Our best results (an equal error rate of 6.3\%)
		  were obtained using NAP and GMM model normalization with
		  the non linear kernel.}
}