Title :
An event-driven f0 weighting for prosody control in a large corpus-based TTS system
Author :
Byeon, Heo-Jin ; Oh, Yung-Hwan
Author_Institution :
Comput. Sci. Div., Korea Adv. Inst. of Sci. & Technol., Daejon, South Korea
Abstract :
This letter presents event-driven weighting methods concentrating on f0 for prosody control in a large corpus-based text-to-speech (TTS) system. We determine the f0 weighting factor for a given target using its linguistic features by automatically using classification and regression trees (CART). The target predicted as perceptually important is weighted more than others. This results in more natural synthetic speech from a prosodic viewpoint.
Keywords :
fault trees; linguistics; regression analysis; speech synthesis; CART; classification and regression tree; event-driven f0 weighting factor; large corpus-based text-to-speech system; linguistic feature; natural synthetic speech; prosodic viewpoint; prosody control; target prediction; Automatic control; Classification tree analysis; Control system synthesis; Control systems; Cost function; Decision trees; Natural languages; Regression tree analysis; Speech synthesis; Weight control;
Journal_Title :
Signal Processing Letters, IEEE
DOI :
10.1109/LSP.2003.821684
