Multistep speaker identification using gibbs-distribution-based extended bayesian inference for rejecting unregistered speaker

Authors:
Yuta Mizobe;Shuichi Kurogi;Tomohiro Tsukazaki;Takeshi Nishida
Affiliations:
Kyushu Institute of Technology, Kitakyushu, Fukuoka, Japan;Kyushu Institute of Technology, Kitakyushu, Fukuoka, Japan;Kyushu Institute of Technology, Kitakyushu, Fukuoka, Japan;Kyushu Institute of Technology, Kitakyushu, Fukuoka, Japan
Venue:
ICONIP'12 Proceedings of the 19th international conference on Neural Information Processing - Volume Part V
Year:
2012

Citing 6
Cited 0

Competitive learning algorithms for vector quantization

Neural Networks
Speaker recognition

Survey of the state of the art in human language technology
Speaker recognition using syllable-based constraints for cepstral frame selection

ICASSP '09 Proceedings of the 2009 IEEE International Conference on Acoustics, Speech and Signal Processing
An analysis of speaker recognition using bagging CAN2 and pole distribution of speech signals

ICONIP'10 Proceedings of the 17th international conference on Neural information processing: theory and algorithms - Volume Part I
Fundamentals of Speaker Recognition

Fundamentals of Speaker Recognition
Naive bayesian multistep speaker recognition using competitive associative nets

ICONIP'11 Proceedings of the 18th international conference on Neural Information Processing - Volume Part I

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a method of multistep speaker identification using Gibbs-distribution-based extended Bayesian inference (GEBI) for rejecting unregistered speaker. The method is developed for our speaker recognition system which utilizes competitive associative nets (CAN2s) for learning piecewise linear approximation of nonlinear speech signal to extract feature vectors of pole distribution from piecewise linear coefficients reflecting nonlinear and time-varying vocal tract of the speaker. In this paper, we focus on the problem of Bayesian inference (BI) in multistep identification for rejecting unregistered speaker and introduce GEBI to solve the problem. The effectiveness of the present method is shown by means of experiments using real speech signals.