Phoneme Generation with Elman-type Neural Networks

DOI

10.2991/wiet-13.2013.6How to use a DOI?

Keywords

Phoneme synthesis; Elman neural networks; Speech processing; Time series analysis; Dynamical systems

Abstract

The paper continues some previous works and analyzes the possibility of phoneme generation with Elman-type neural networks. The vowels are composed of repetitions of the same elemental pattern with certain variations (assumed to be chaotic) of the signal parameters. The Elman-type network proved adequate to learn and generate independently in an open loop the elementals with signal parameters controlled by chaotic sources. The experimental results demonstrated that the technique can be further developed and used in synthesized speech.

Copyright

© 2013, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Marius Crisan
PY  - 2013/12
DA  - 2013/12
TI  - Phoneme Generation with Elman-type Neural Networks
BT  - Proceedings of the AASRI Winter International Conference on Engineering and Technology (AASRI-WIET 2013)
PB  - Atlantis Press
SP  - 23
EP  - 26
SN  - 1951-6851
UR  - https://doi.org/10.2991/wiet-13.2013.6
DO  - 10.2991/wiet-13.2013.6
ID  - Crisan2013/12
ER  -