Fuzzy voice classifier and fuzzy voice pathology identification system

Abstract

Speech is one of the common modes of communication and it is a process of transferring information from one entity to another. In recent years there has been much research on unvoiced/voiced classification and voice pathology classification. In this research work a simple fuzzy classifier has been designed to segment the voiced and unvoiced portions of a speech signal. A simple feature extraction algorithm is proposed to extract the Tri Mean relative average perturbation (Tri Mean-RAP) features from the segmented voice portion of the speech signal. Further, the extracted Tri Mean-RAP features are used to model the fuzzy pathology identification system. In the proposed fuzzy classifier, the energy, change in energy and difference in the change in energy level between the adjacent frames are fuzzified and rules are formulated to segment the voiced portion. The Tri Mean-RAP features are then extracted from the segmented voice portion, the Tri Mean-RAP features are fuzzified and rules are formulated to identify the voice pathology of the speech signal. The proposed methods are validated through simulation.