Stochastic Pitch Prediction Improves the Diversity and Naturalness of Speech in Glow-TTS.
Sewade OgunVincent ColotteEmmanuel VincentPublished in: CoRR (2023)
Keyphrases
- text to speech
- speech synthesis
- linear prediction
- prosodic features
- prediction accuracy
- finite state transducers
- speech recognition
- formant frequencies
- speech signal
- prediction algorithm
- fundamental frequency
- prediction error
- acoustic features
- prediction model
- stochastic model
- emotion recognition
- speaker identification
- spoken language
- neural network
- dialogue system
- human computer interaction
- reinforcement learning
- decision trees