Unsupervised Acoustic Unit Discovery for Speech Synthesis Using Discrete Latent-Variable Neural Networks.
Ryan EloffAndré NortjeBenjamin van NiekerkAvashna GovenderLeanne NortjeArnu PretoriusElan Van BiljonEwald van der WesthuizenLisa van StadenHerman KamperPublished in: INTERSPEECH (2019)
Keyphrases
- speech synthesis
- latent variables
- prosodic features
- neural network
- global exponential stability
- speech recognition
- probabilistic model
- text to speech
- latent variable models
- prior knowledge
- real valued
- vocal tract
- hierarchical model
- random variables
- pattern recognition
- topic models
- observational data
- unsupervised learning
- gaussian process
- semi supervised
- topic modeling
- back propagation
- hidden variables
- supervised learning
- knowledge discovery
- hopfield neural network
- language model
- observed variables
- probabilistic latent semantic analysis
- multilayer perceptron
- hidden markov models