UNFOLD: a memory-efficient speech recognizer using on-the-fly WFST composition.

Reza Yazdani José-María Arnau Antonio González

Published in: MICRO (2017)

Keyphrases

memory efficient
speech recognizer
speech recognition
spoken language
speech understanding
speech recognizers
broadcast news
speech recognition systems
hidden markov models
neural network
bayesian networks
language model
automatic speech recognition
multi modal