Fastmandarin: Efficient Local Modeling for Natural Mandarin Speech Synthesis.

Chenglong Jiang Ying Gao Hao Jin Linrong Pan Wing W. Y. Ng

Published in: ICASSP (2024)

Keyphrases

speech synthesis
speech recognition
prosodic features
real world
decision trees
pattern recognition
lightweight
cost effective
computationally expensive
real time
data sets
multiscale
data structure
artificial neural networks
non stationary
vocal tract