Corpus ID: 142030535

Surveillance de réseaux professionnels de communication par reconnaissance du locuteur

@inproceedings{Preti2008SurveillanceDR,
  title={Surveillance de r{\'e}seaux professionnels de communication par reconnaissance du locuteur},
  author={A. Preti},
  year={2008}
}
Ce travail de these s'interesse a la reconnaissance automatique du locuteur (RAL) dans les reseaux professionnels de communication (Private Mobile Radio networks : PMR). Plus precisement, nous nous interessons a la surveillance des utilisateurs en cours de communication pour detecter un changement de locuteur, issu du vol ou du pret d'un terminal de communication. Les systemes « etat de l'art » de RAL presentent aujourd'hui de tres bonnes performances sur des signaux de conversations… Expand
Close Speakers Model and Comparative Study in Automatic Speaker Verification
The performance of speaker verification system degrades when the test segments are utterances of short duration, therefore, we investigate the use of model representing our target speaker with hisExpand
Optimal impostor model in automatic speaker verification
  • H. Djellali, M. Laskri
  • Computer Science
  • 2012 IEEE International Conference on Complex Systems (ICCS)
  • 2012
TLDR
The present study demonstrates that several codebooks for Universal Background Models give better results and the VQ Partial method achieved less half total error rate for 128 codebook size better than Baseline Vector Quantization approach for 32 codebook sizes. Expand
Clustering Algorithm in Automatic Speaker Verification
We propose a new modeling approach in Automatic Speaker Verification A.S.V based on Gaussians Mixtures Models and Maximum a posteriori adaptation MAP. We propose clustering algorithm for intra andExpand
Using Vector Quantization for Universal Background Model in Automatic Speaker Verification
TLDR
The present study demonstrates that the multiples codebook gives more verification accuracy called equal error rate but this improvement also depends on the codebook size. Expand
Using vector quantization in Automatic Speaker Verification
TLDR
Several technique based on vector quantization and maximum a posteriori adaptation in Automatic Speaker Verification ASV are investigated and compared with traditional approach in VQ, MAP adaptation and Gaussian Mixture Models. Expand