34 search results - page 7 / 7 » Mel Frequency Cepstral Coefficients for Music Modeling |
SPEECH
2011
12 years 11 months ago
2011
We propose a novel algorithm for the separation of convolutive speech mixtures using two-microphone recordings, based on the combination of independent component analysis (ICA) an...
TASLP
2008
13 years 4 months ago
2008
Abstract--We describe in this paper an audio denoising technique based on sparse linear regression with structured priors. The noisy signal is decomposed as a linear combination of...
TASLP
2010
12 years 11 months ago
2010
In the presence of environmental noise, speakers tend to adjust their speech production in an effort to preserve intelligible communication. The noise-induced speech adjustments, c...
ICASSP
2010
IEEE
13 years 4 months ago
2010
IEEE
This paper considers the enhancement of noisy speech. Earlier studies have revealed that an approach that enhances spectral envelopes by using prior knowledge about the all-pole (...