PARALLEL MODEL COMBINATION FOR SPEECH RECOGNITION IN ADDITIVE AND CONVOLUTIONAL NOISE
M. J. F. Gales and S. J. Young
December 1993
This paper addresses the problem of speech recognition in the presence of both additive and convolutional noise. A new scheme is described, which is a simple extension to the standard Parallel Model Combination (PMC) technique. A modified `mismatch' function is introduced which accounts for the effects of convolutional noise. This `mismatch' function is then used to estimate the difference in channel conditions between training and test environments. Having estimated the tilt parameters, Maximum Likelihood (ML) estimates of the corrupted speech model may be obtained. The scheme is evaluated using the NOISEX-92 database. The performance in the presence of both interfering additive noise and convolutional noise shows only slight degradation compared with that obtained when no convolutional noise is present.
If you have difficulty viewing files that end '.gz',
which are gzip compressed, then you may be able to find
tools to uncompress them at the gzip
web site.
If you have difficulty viewing files that are in PostScript, (ending
'.ps' or '.ps.gz'), then you may be able to
find tools to view them at
the gsview
web site.
We have attempted to provide automatically generated PDF copies of documents for which only PostScript versions have previously been available. These are clearly marked in the database - due to the nature of the automatic conversion process, they are likely to be badly aliased when viewed at default resolution on screen by acroread.