Learning noise robust ResNet-based speaker embedding for speaker recognition - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2022

Learning noise robust ResNet-based speaker embedding for speaker recognition

Mohammad Mohammadamini
  • Fonction : Auteur
  • PersonId : 1133019
Driss Matrouf
  • Fonction : Auteur
  • PersonId : 1133020
Jean-François Bonastre
  • Fonction : Auteur
Sandipana Dowerah
  • Fonction : Auteur
Romain Serizel
  • Fonction : Auteur
  • PersonId : 1125745
Denis Jouvet

Résumé

The presence of background noise and reverberation, especially in far distance speech utterances diminishes the performance of speaker recognition systems. This challenge is addressed on different levels from the signal level in the front end to the scoring technique adaptation in the back end. In this paper, two new variants of ResNet-based speaker recognition systems are proposed that make the speaker embedding more robust against additive noise and reverberation. The goal of the proposed systems is to extract x-vectors in noisy environments that are close to their corresponding x-vector in a clean environment. To do so, the speaker embedding network minimizes the speaker classification loss function and the distance between pairs of noisy and clean x-vectors jointly. The experimental results obtained by our systems are compared with the baseline ResNet system. In different situations with real and simulated noises and reverberation conditions, the modified systems outperform the baseline ResNet system. The proposed systems are tested with four evaluation protocols. In the presence of artificial noise and reverberation, we achieved 19% improvement of EER. The main advantage of the proposed systems is their efficiency against real noise and reverberation. In the presence of real noise and reverberation, we achieved 15% improvement of EER.
Fichier principal
Vignette du fichier
Learning noise robust ResNet-based speaker embedding for speaker recognition.pdf (192.9 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-03650549 , version 1 (25-04-2022)

Identifiants

  • HAL Id : hal-03650549 , version 1

Citer

Mohammad Mohammadamini, Driss Matrouf, Jean-François Bonastre, Sandipana Dowerah, Romain Serizel, et al.. Learning noise robust ResNet-based speaker embedding for speaker recognition. Odyssey 2022 : The Speaker and Language Recognition Workshop, Jun 2022, Beijing, China. ⟨hal-03650549⟩

Collections

UNIV-AVIGNON LIA
258 Consultations
458 Téléchargements

Partager

Gmail Facebook X LinkedIn More