A method of creating a new speaker²s voicefont in a text-to-speech system.

Takashi Saito Masaharu Sakamoto

Published in: INTERSPEECH (2000)

Keyphrases

dynamic programming
text to speech
high precision
synthetic data
significant improvement
detection method
experimental evaluation
computational complexity
high accuracy
input data
model selection
clustering method
cost function
pairwise
probabilistic model
prior knowledge
similarity measure