Speech Signal Analysis and Classification of Dominant Parameter for Pathological Voices

Abstract

The primary objective of the project is to analyze speech signals by determining the important parameters that affect the voice of an individual which leads to various voice disorders. The analysis is carried out based on the individual’s age and gender with the help of the pattern recognized from each sample and the value of each parameter is compared with the nominal values of the healthy person with respect to their age and gender using the Praat software. The secondary objective is the classification of the voice signal into normal and abnormal voice samples using the machine learning software Konstanz Information Miner (KNIME).