On Developing An Automatic Speech Recognition System For Standard Arabic Language

The Automatic Speech Recognition (ASR) applied to Arabic language is a challenging task. This is mainly related to the language specificities which make the researchers facing multiple difficulties such as the insufficient linguistic resources and the very limited number of available transcribed Arabic speech corpora. In this paper, we are interested in the development of a HMM-based ASR system for Standard Arabic (SA) language. Our fundamental research goal is to select the most appropriate acoustic parameters describing each audio frame, acoustic models and speech recognition unit. To achieve this purpose, we analyze the effect of varying frame windowing (size and period), acoustic parameter number resulting from features extraction methods traditionally used in ASR, speech recognition unit, Gaussian number per HMM state and number of embedded re-estimations of the Baum-Welch Algorithm. To evaluate the proposed ASR system, a multi-speaker SA connected-digits corpus is collected, transcribed and used throughout all experiments. A further evaluation is conducted on a speaker-independent continue SA speech corpus. The phonemes recognition rate is 94.02% which is relatively high when comparing it with another ASR system evaluated on the same corpus. {"references": ["M. Kabache, and M. Guerti, \"Application des r\u00e9seaux de neurones \u251c\u00e1 la\nreconnaissance des phon\u00e8mes sp\u00e9cifiques \u251c\u00e1 l-Arabe standard\", SETIT\n2005, Tunisia, March 2005.", "S.A. Selouani, and J. Caelen, \"Recognition of phonetic features using\nneural networks and knowledge-based system: a comparative study\",\nInternational Journal on artificial intelligence tools, world scientific\npublishing editors, vol. 8, no. 1, pp. 73-103, 1999.", "S. Hazmoune, F. Bougamouza, and M. Benmohammed, \"La\nreconnaissance automatique de la parole par combinaison de classifieurs\nmarkoviens\", JEESI-09, 2009.", "Y. A. Alotaibi, \"Comparative Study of ANN and HMM to Arabic Digits\nRecognition Systems\", JKAU: Eng. Sci., vol. 19, no. 1, pp. 43-60, 2008.", "R. Ejbali, Y. Ben Ayed, and A. M. Alimi, \"Arabic continues speech\nrecognition system using context-independent\", Sixth International\nMulti-Conference on Systems, Signals & Devices, Tunisia, March 2009.", "M. Elshafei, \"Toward an Arabic text-to-speech system,\" The Arabian\nJournal for Science and Engineering, vol. 16, no. 4, pp. 565-583, 1991.", "M. Alkhouli, Linguistic Phonetics, Daar Alfalah, Swaileh, Jordan, 1990.", "F. Jelinek, \"Continuous speech recognition by statistical methods\",\nProc. IEEE, vol. 64, pp. 532-556, 1976.", "L. R. Rabiner, \"A Tutorial on Hidden Markov Models and selected\napplications in Speech Recognition\", Proc. IEEE, vol. 77, pp. 257-286,\nFeb. 1989.\n[10] F. Lefevre, Estimation de probabilit\u00e9 non-param\u00e9trique pour la\nreconnaissance Markovienne de la parole, Pierre and Marie Curie\nUniversity, Jan. 2000.\n[11] S. Young et al., The HTK Book (for HTK Version 3.4), Cambridge\nUniversity, March 2009.\n[12] M. Boudraa, and B. Boudraa \"Twenty list of ten arabic Sentences for\nAssessment\", ACUSTICA acta acoustica. vol. 86, no. 43.71, pp. 870-\n882, Nov. 1998."]}