Machine learning for Arabic phonemes recognition using electrolarynx speech

<p><span lang="EN-US">Automatic speech recognition system is one of the essential ways interaction with machines. Interests in based intelligent systems have grown past few decades. Therefore, there a need to develop more efficient methods for human ensure reliability communication between individuals and This paper concerned Arabic phoneme electrolarynx device. Electrolarynx device used by cancer patients having vocal laryngeal cords removed. Speech here considered find preferred machine learning model that can classify phonemes produced The employs different schemes, including convolutional neural network, recurrent artificial network (ANN), random forest, extreme gradient boosting (XGBoost), long short-term memory. Modern standard utilized testing training phases system. dataset covers both an ordinary recorded same person. Mel frequency cepstral coefficients are as features. results show ANN method outperformed other accuracy rate 75%, precision value 77%, error (PER) 21.85%.</span></p>