Explicit duration modelling in HMM-based speech synthesis using a hybrid hidden Markov model-multilayer perceptron.
Kalu U. OgburekeJoão P. CabralJulie Carson-BerndsenPublished in: SAPA@INTERSPEECH (2012)
Keyphrases
- hidden markov models
- multilayer perceptron
- speech synthesis
- speech recognition
- hidden semi markov models
- neural network
- variable duration
- artificial neural networks
- back propagation
- vocal tract
- automatic speech recognition
- radial basis function
- speech signal
- dynamic probabilistic networks
- sequential data
- neural network model
- gesture recognition
- markov models
- viterbi algorithm
- noisy environments
- speech corpus
- conditional random fields
- text to speech
- speaker identification
- speaker verification
- handwriting recognition
- markov model
- continuous hidden markov models
- maximum likelihood
- language model
- feature vectors
- hierarchical hidden markov model