Generative Acoustic-Phonemic-Speaker Model Based on Three-Way Restricted Boltzmann Machine.

Toru Nakashika Yasuhiro Minami

Published in: INTERSPEECH (2016)

Keyphrases

restricted boltzmann machine
speech sounds
deep learning
acoustic features
automatic speech recognition
generative model
speech signal
unsupervised learning
speech recognition
conditional random fields
probabilistic graphical models
vocal tract
learning algorithm
active learning