Application of triphone clustering in acoustic modeling for continuous speech recognition in Bengali

The performance of the acoustic models is highly reflective on the overall performance of any continuous speech recognition system. Hence generation of an accurate and robust acoustic model holds the key to satisfactory recognition performance. As phones are found to vary according to the position of occurrence within a particular word, context information is of prime importance in acoustic modeling of phonetic signals. In this paper we look at the effect of triphone-based acoustic modeling over monophone based acoustic models in the context of continuous speech recognition in Bengali. Keeping in mind the lack of training resources for triphone-based acoustic modeling in Bengali, we have also described herein, the method of generating triphone clusters using decision tree based techniques. These triphone clusters have then been used to generate tied-state triphone based acoustic models to be used in a continuous speech recognizer.

1. Introduction

Continuous speech recognition has been an area of active research for quite some time now. However when compared to languages like English or French, the state of speech research involving Indian languages is yet to gain momentum. Although some amount of effective research has gone into the development of speech recognizers in Hindi [1] and some south Indian Languages [2], the research scenario for Bengali language is far from satisfactory. In course of our effort to create a continuous speech recognizer for Bengali, we came across certain issues involving the generation of robust acoustic models for Bengali.