Skoči na glavni sadržaj

International Journal for Engineering Modelling, Vol. 20 No. 1-4, 2007.

Izvorni znanstveni članak

Recognition of speaking in a phone channel with a neural network backpropagation

Luis A. Cruz-Beltrán ; Seccion de Estudios de Postgrado e Investigation, Escuela Superior de Ingenieria Mecanica y Electrica, Instituto Politécnico Nacional, Edif. Z-4, 3er. Piso, Col. Lindavista, C.P. 07738, MEXICO D.F.
Marco Antonio Acevedo-Mosqueda ; Seccion de Estudios de Postgrado e Investigation, Escuela Superior de Ingenieria Mecanica y Electrica, Instituto Politécnico Nacional, Edif. Z-4, 3er. Piso, Col. Lindavista, C.P. 07738, MEXICO D.F.
Jose Luis López-Bonilla ; Seccion de Estudios de Postgrado e Investigation, Escuela Superior de Ingenieria Mecanica y Electrica, Instituto Politécnico Nacional, Edif. Z-4, 3er. Piso, Col. Lindavista, C.P. 07738, MEXICO D.F.

Puni tekst: engleski pdf 136 Kb

str. 17-22

preuzimanja: 93

citiraj

Sažetak

In this paper we propose an algorithm for identification of speaking in the phone channel. The investigation is based on the behavior of the Artificial Neural Network (ANN) Backpropagation. The analysed voice is captured with format *.wav. We introduce the procedure for the speaking identification in the phone channel which offered the best results in several tests that we carry out with our method.

Ključne riječi

artificial neural networks; wavelets; backpropagation; CLP; wav

Hrčak ID:

187454

URI

https://hrcak.srce.hr/187454

Datum izdavanja:

27.1.2009.

Posjeta: 330 *