METHODOLOGY OF DEVELOPMENT OF AN ARABIC-BASED SPEECH PROCESSOR | ||||
International Conference on Aerospace Sciences and Aviation Technology | ||||
Article 49, Volume 2, A.S.A.T. CONFERENCE 21-23 April 1987 , CAIRO, April 1987, Page 1119-1134 PDF (3.1 MB) | ||||
Document Type: Original Article | ||||
DOI: 10.21608/asat.1987.26211 | ||||
View on SCiNiTO | ||||
Author | ||||
N. M. El-Nadi | ||||
Brig.Gen.Dr.Eng., The author is currently the Chief of Chair of EW in the M.T.C. | ||||
Abstract | ||||
The paper presents a methodology of developing a Text-to-Speech Arabic synthesizer based on the linear independent speech production model shown in Fig.(1) Relevent phonetical properties of Arabic are covered less the ambiguities ever associated in most references dealing with Arabic Phonology /1/,/2/,/3/,/4/,/5/&/6/. Such ambiguities have always posed difficult problems that delayed the production of quality synthetic Arabic speech. To the best of the authors knowlege no Arabic voice processor of quality speech exists. The paper describes how to apply state-of -the art techniques of digital signal processing for the estimation of the parameters of the speech synthesizer model for the production of Arabic Utterances. Numerous sonograms(time-frequency diagrams) for Arabic utterances are presented,which illustrate the relevence of the methodological preprocessing neede . The author presents a proposed model for text-to-speech synthesizer model favouring the concatenation of VCV (Vowel+Consonant+Vowel) segments rather than phonemes or allophones.This model will include the transitions from a vowel to a consonant and from a consonant to a vowel in a natural form and promises an improve-ment in performance. | ||||
Statistics Article View: 82 PDF Download: 257 |
||||