Anchor and UBM-based Multi-Class MLLR M-Vector System for Speaker Verification - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2013

Anchor and UBM-based Multi-Class MLLR M-Vector System for Speaker Verification

Résumé

In this paper, we propose two techniques to extend the recently introduced global Maximum Likelihood Linear Regression (MLLR) transformation (i.e. super-vector) based m-vector system for speaker verification into a multi-class MLLR m-vector system in the Universal Background Model (UBM) framework. In the first method, Gaussian mean vectors of the UBM are first grouped into several classes using conventional K-means and a proposed clustering algorithm based on Expectation Maximization (EM) and Maximum Likelihood (ML) concepts. Then, MLLR transformations are calculated for a given speech data with respect to each class, which are used in the form of super-vector for speaker representation by their m-vectors. In the second approach, several MLLR transformations are estimated with respect to pre-defined models called anchors. The proposed systems show better performance than the conventional system. Furthermore, the proposed UBM-based system does not require additional alignment of speech data with respect to the UBM for estimation of multiple MLLR transformations. We also further show that the proposed EM & ML clustering algorithm is robust to random initialization and provides equal or comparable system performance compared to K-means. The experimental results are shown on NIST 2008 SRE core condition over various tasks.
Fichier principal
Vignette du fichier
i13_2450.pdf (161.54 Ko) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte
Loading...

Dates et versions

hal-01690250 , version 1 (23-01-2018)

Identifiants

  • HAL Id : hal-01690250 , version 1

Citer

Achintya K Sarkar, Claude Barras. Anchor and UBM-based Multi-Class MLLR M-Vector System for Speaker Verification. Interspeech 2013, Aug 2013, Lyon, France. ⟨hal-01690250⟩
26 Consultations
44 Téléchargements

Partager

Gmail Facebook X LinkedIn More