Title:
|
HMM-based Spanish speech synthesis using CBR as F0 estimator
|
Author:
|
Gonzalvo Fructuoso, Xavier; Iriondo Sanz, Ignasi; Socoró Carrié, Joan Claudi; Alías Pujol, Francesc; Monzo Sánchez, Carlos
|
Other authors:
|
Universitat Ramon Llull. La Salle |
Abstract:
|
Hidden Markov Models based text-to-speech (HMM-TTS) syn thesis is a technique for generating speech from trained statisti cal models where spectrum, pitch and durations of basic speech
units are modelled altogether. The aim of this work is to de scribe a Spanish HMM-TTS system using CBR as a F0 esti mator, analysing its performance objectively and subjectively.
The experiments have been conducted on a reliable labelled
speech corpus, whose units have been clustered using contex tual factors according to the Spanish language. The results
show that the CBR-based F0 estimation is capable of improving
the HMM-based baseline performance when synthesizing non declarative short sentences and reduced contextual information
is available. |
Publication date:
|
2007-05 |
Subject (UDC):
|
62 - Enginyeria. Tecnologia |
Subject(s):
|
Processament de la parla Anàlisi prosòdica (Lingüística) |
Rights:
|
© International Speech Communication Association. Tots els drets reservats |
Pages:
|
4 p. |
Document type:
|
Conference Object |
Published by:
|
ITRW on Nonlinear Speech Processing, Paris, 22-25 of May 2007
|
Share:
|
|