PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Mask estimation for missing data speech recognition based on statistics of binaural interaction
S. Harding, J.P. Barker and G.J. Brown
IEEE Trans Speech and Audio Processing 2005.

Abstract

This paper describes a perceptually motivated computational auditory scene analysis (CASA) system that combines sound separation according to spatial location with the `missing data' approach for robust speech recognition in noise. Missing data time-frequency masks are created using probability distributions based on estimates of interaural time and level differences (ITD and ILD) for mixed utterances in reverberated conditions; these masks indicate which regions of the spectrum constitute reliable evidence of the target speech signal. A number of experiments compare the relative ef cacy of the binaural cues when used individually and in combination. We also investigate the ability of the system to generalise to acoustic conditions not encountered during training. Performance on a continuous digit recognition task using this method is found to be good, even in a particularly challenging environment with three concurrent male talkers.

PDF - PASCAL Members only - Requires Adobe Acrobat Reader or other PDF viewer.
EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
Speech
ID Code:1898
Deposited By:Jon Barker
Deposited On:29 December 2005