Sciweavers

Share
IJON
2008

Discovering speech phones using convolutive non-negative matrix factorisation with a sparseness constraint

8 years 6 months ago
Discovering speech phones using convolutive non-negative matrix factorisation with a sparseness constraint
Discovering a representation that allows auditory data to be parsimoniously represented is useful for many machine learning and signal processing tasks. Such a representation can be constructed by Non-negative Matrix Factorisation (NMF), a method for finding parts-based representations of non-negative data. Here, we present an extension to convolutive NMF that includes a sparseness constraint, where the resultant algorithm has multiplicative updates and utilises the beta divergence as its reconstruction objective. In combination with a spectral magnitude transform of speech, this method discovers auditory objects that resemble speech phones along with their associated sparse activation patterns. We use these in a supervised separation scheme for monophonic mixtures, finding improved separation performance in comparison to classic convolutive NMF.
Paul D. O'Grady, Barak A. Pearlmutter
Added 12 Dec 2010
Updated 12 Dec 2010
Type Journal
Year 2008
Where IJON
Authors Paul D. O'Grady, Barak A. Pearlmutter
Comments (0)
books