Linguistic and Mixed Excitation Improvements on a HMM-based speech synthesis for Castilian Spanish

Xavier Gonzalvo, Joan Claudi Socoró, Ignasi Iriondo, Carlos Monzo, Elisa Martínez

Producción científica: Contribución a una conferenciaContribuciónrevisión exhaustiva

13 Citas (Scopus)

Resumen

Hidden Markov Models based text-to-speech(HMM-TTS) synthesis is one of the techniques for generating speech from trained statistical models where spectrum and prosody of basic speech units are modelled altogether. This paper presents the advances in our Spanish HMM-TTS and a perceptual test is conducted to compare it with an extended PSOLA-based concatenative (E-PSOLA) system. The improvements have been performed on phonetic information and contextual factors according to the Castilian Spanish language and speech generation using a mixed excitation (ME) technique. The results show the preference of the new HMM-TTS system in front of the previous system and a better MOS in comparison with a real E-PSOLA in terms of acceptability, intelligibility and stability.

Idioma originalInglés
Páginas362-367
Número de páginas6
EstadoPublicada - 2007
Evento6th ISCA Tutorial and Research Workshop on Speech Synthesis, SSW 2007 - Bonn, Alemania
Duración: 22 ago 200724 ago 2007

Conferencia

Conferencia6th ISCA Tutorial and Research Workshop on Speech Synthesis, SSW 2007
País/TerritorioAlemania
CiudadBonn
Período22/08/0724/08/07

Huella

Profundice en los temas de investigación de 'Linguistic and Mixed Excitation Improvements on a HMM-based speech synthesis for Castilian Spanish'. En conjunto forman una huella única.

Citar esto