Hierarchical Multi-Grained Generative Model for Expressive Speech Synthesis.
Yukiya HonoKazuna TsuboiKei SawadaKei HashimotoKeiichiro OuraYoshihiko NankakuKeiichi TokudaPublished in: INTERSPEECH (2020)
Keyphrases
- generative model
- speech synthesis
- speech recognition
- pitman yor process
- probabilistic model
- text to speech
- mixture model
- bayesian framework
- prior knowledge
- discriminative learning
- prosodic features
- vocal tract
- em algorithm
- posterior probability
- hierarchical bayesian model
- semi supervised
- language model
- latent dirichlet allocation
- discriminative models
- generative and discriminative models
- expectation maximization
- machine learning
- neural network