Abstract:

Literature counts several algorithms whose aim is to detect vocal pathologies contained in the acoustic signal. The majority of these works are limited to classify the produced speech to healthy or pathological. The goal of this work is to develop an advanced algorithm allowing the classification of vocal pathologies. In this paper, we distinguish three types of vocal disabilities: the addition, the suppression and the substitution. This classification is based on the comparison between the referenced phonetic model of the spoken Arabic language and that proper to the concerned speaker. Our proposition reached a classification rate of 86%. Speech therapists, biologists and computer scientists can benefit from this work to automate its diagnostics.