ITERATIVE GROUP SELECTION-BASED ENHANCEMENT OF TIME-FREQUENCY MASKS FOR MISSING DATA RECOGNITION

Abstract

Missing data approaches have recently been applied to speech recognition tasks to increase noise robustness. The drawback of missing data techniques is the vulnerability of the recognizer to errors in the reliability mask. This work proposes a novel group selection algorithm to perform top-down refinement of initial bottom-up reliability mask estimates with the goal of removing these errors. A novel probabilistic decision process based on normalized likelihood distances is proposed and used to evaluate the quality of a reliability mask without any a priori noise knowledge. Experimental results on a speaker identification task illustrate the ability of the combined bottom-up top-down system to significantly outperform traditional bottom-up only missing data techniques for various types of mask corruption.