Speech parameter generation algorithms for HMM-based speech synthesis.
Keiichi TokudaTakayoshi YoshimuraTakashi MasukoTakao KobayashiTadashi KitamuraPublished in: ICASSP (2000)
Keyphrases
- speech synthesis
- text to speech
- speech recognition
- recently developed
- optimal parameters
- neural network
- data structure
- computational complexity
- stopping criterion
- optimization problems
- computationally efficient
- computational efficiency
- vocal tract
- parameter tuning
- image acquisition
- worst case
- hidden markov models
- significant improvement
- bayesian networks
- feature extraction
- learning algorithm