A Deep Generative Model of Vowel Formant Typology.
Ryan CotterellJason EisnerPublished in: NAACL-HLT (2018)
Keyphrases
- generative model
- speech signal
- vocal tract
- probabilistic model
- speech recognition
- posterior probability
- em algorithm
- bayesian framework
- mixture model
- speech synthesis
- discriminative models
- latent dirichlet allocation
- prior knowledge
- discriminative learning
- expectation maximization
- topic models
- automatic speech recognition
- speaker independent
- language model
- semi supervised
- dirichlet process mixture models
- conditional random fields
- generative and discriminative models
- primal sketch
- image processing
- pairwise
- multiscale