Speech recognition in noisy reverberant rooms using a frequency domain blind deconvolution method.

George Nokas Evangelos Dermatas

Published in: EUROSPEECH (1999)

Keyphrases

frequency domain
speech recognition
blind deconvolution
cross correlation
power spectrum
speech signal
similarity measure
language model
spatial domain
noisy environments
computational complexity
hidden markov models
feature set
generative model
additive noise