Automatic segmentation of Arabic speech signals by HMM and ANN

In this paper, we propose an automatic segmentation system of speech into phonemes for the Arabic language. This segmentation is based on two different techniques : Hidden Markov Models (HMM) and Artificial Neural Networks (ANN). Both systems were used to classify the speech signals, extracted from ALGASD corpus (ALGerian Arabic Speech Database), into five classes : fricatives, plosives, nasals, liquids and vowels. These methods achieve important performances with advantage of ANN.