PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

A max kernel for text-independent speaker verification systems
Johnny Mariéthoz and Samy Bengio
In: Second Workshop on Multimodal User Authentication, MMUA(2006).

Abstract

In this paper, we present a principled SVM based speaker verification system. A general approach is developed that enables the use of any kernel at the frame level. An extension of his approach using the Max operator is then proposed. The new system is then compared to state-of-the-art GMM and other SVM based systems found in the literature on the Polyvar database. It is found that the new system outperforms, most of the time, the other systems, statistically significantly.

PDF - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Conference or Workshop Item (Oral)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Speech
ID Code:2491
Deposited By:Samy Bengio
Deposited On:22 November 2006