A method to extract articulatory parameters from the speech signal using neural networks

Author

Branco, Antonio ; Tomé, Ana ; Teixeira, Antonio ; Vaz, Francisco

Author_Institution

Dept. de Electron. e Telecoms, Aveiro Univ., Portugal

Volume

2

fYear

1997

fDate

2-4 Jul 1997

Firstpage

583

Abstract

We present a method that uses artificial neural networks for acoustic to articulatory mapping. An assembly of Kohonen (1982) neural nets is used, in the first stage a network maps cepstral values, each neuron contains a subnet in a second stage that maps the articulatory space. The method allows both the acoustic to articulatory mapping, ensuring smooth varying vocal tract shapes, and the study of the nonuniqueness problem

Keywords

acoustic signal processing; cepstral analysis; feature extraction; learning (artificial intelligence); linear predictive coding; self-organising feature maps; speech coding; speech synthesis; Kohonen neural nets; LPC derived cepstral parameters; acoustic to articulatory mapping; articulatory parameters extraction; articulatory space; artificial neural networks; cepstral values; neural networks training; nonuniqueness problem; smooth varying vocal tract shapes; speech signal; speech synthesis; Assembly; Cepstral analysis; Linear predictive coding; Lips; Neural networks; Neurons; Shape; Signal mapping; Speech; Tongue;

fLanguage

English

Publisher

ieee

Conference_Titel

Digital Signal Processing Proceedings, 1997. DSP 97., 1997 13th International Conference on

Conference_Location

Santorini

Print_ISBN

0-7803-4137-6

Type

conf

DOI

10.1109/ICDSP.1997.628417

Filename

628417