Voice biometric system: The identification of the severity of Cerebral Palsy using mel-frequencies stochastics approach

Abstract

Cerebral Palsy (CP) is a neurological condition that causes problems in body movement and muscle control that can inhibit the development of children's speech. It can be classified into several types, based on the stiffness of the muscles that they suffer. This study offers a new technique for identifying the severity CP level of children based on their speech. This study utilizes the capabilities of the voice biometrics system (VBS) to authenticate people based on their voice. The Mel-frequency stochastic model is also offered as a new approach in the feature extraction process. Because of the pattern of speech signals of children with CP which is irregular, then neuro fuzzy is chosen as a method in the speech classifier. Based on the experiment conducted to respondents, the accuracy of the technique is 87.5%. This result shows good performance of the new approach for realizing the research objective.