5th International Conference on Spoken Language Processing

Sydney, Australia
November 30 - December 4, 1998

Speaker Recognition Using Residual Signal Of Linear and Nonlinear Prediction Models

Marcos Faundez-Zanuy (1), Daniel Rodriguez-Porcheron (2)

(1) Escola Universitaria Politecnica de Mataro, Spain
(2) Universidad Politecnica de Catalunya, Spain

This Paper discusses the usefullness of the residual signal for speaker recognition. It is shown that the combination of both a measure defined over LPCC coefficients and a measure defined over the energy of the residual signal gives rise to an improvement over the classical method which considers only the LPCC coefficients. If the residual signal is obtained from a linear prediction analisys, the improvement is 2.63% (error rate drops from 6.31% to 3.68%) and if it is computed through a nonlinear predictive neural nets based model, the improvement is 3.68%.

Full Paper

Bibliographic reference.  Faundez-Zanuy, Marcos / Rodriguez-Porcheron, Daniel (1998): "Speaker recognition using residual signal of linear and nonlinear prediction models", In ICSLP-1998, paper 1102.