Generation of creaky voice for improving the quality of HMM-based speech synthesis.

N. P. Narendra K. Sreenivasa Rao

Published in: Comput. Speech Lang. (2017)

Keyphrases

speech synthesis
text to speech
speech recognition
hidden markov models
prosodic features
high quality
vocal tract
quality assessment
higher quality
data quality
neural network
improve quality
speech corpus
user satisfaction