Data duration and model size dependency of GMM- and SVM-based speaker verification performance GKM ve DVM tabanli konuşmaci doǧrulamada veri süresi ve model büyüklüǧüne dayali başarim analizi


Dikici E., SARAÇLAR M.

2009 IEEE 17th Signal Processing and Communications Applications Conference, SIU 2009, Antalya, Türkiye, 9 - 11 Nisan 2009, ss.337-340, (Tam Metin Bildiri)

  • Yayın Türü: Bildiri / Tam Metin Bildiri
  • Doi Numarası: 10.1109/siu.2009.5136401
  • Basıldığı Şehir: Antalya
  • Basıldığı Ülke: Türkiye
  • Sayfa Sayıları: ss.337-340
  • Boğaziçi Üniversitesi Adresli: Evet

Özet

This study aims to present two methods for speaker verification problem using Gaussian Mixture Models and Support Vector Machines over the CSLU Speaker Recognition Database. The performance of these two classifiers for various training and test data durations and model sizes are compared. The effect of score fusion on verification error rates is investigated. ©2009 IEEE.