Gender factor suppression for Bangla ASR

Abstract

Hidden factor such as gender characteristic plays an important role on the performance of Bangla (widely used as Bengali) automatic speech recognition (ASR). If there is a suppression process that represses the decrease of differences in acoustic-likelihood among categories resulted from gender factors, a robust ASR system can be realized. In our previous paper, we proposed a technique for gender effects suppression that composed two hidden Markov model (HMM)-based classifiers that focused on the gender factor. In the current study, we utilize three HMM-based classifiers corresponding to male, female and gender-independent (GI) characteristics. In an experiment on a collected Bangla speech database, the proposed system that incorporates GI-classifier has achieved a significant improvement over prior art in terms of both word and sentence accuracy. Moreover, the proposed system requires fewer mixture components in HMM and hence, reduces computation time.