Pitch Contours in Affirmative and Interrogative Arabic Sentences

Abstract

Over the years, men have been trying to generate or simulate the human voice (speech). They first started by simulating some vowels, then they evolved to generate full words or short sentences. The modem word used for this process is 'Speech Synthesis'.

Based on the well known principle 'First things first', researchers concentrated on obtaining the highest possible graphem-to-phoneme translation fidelity leading to minimum ambiguity in the synthesized speech. Listeners to the output of modem speech synthesizers will no longer confuse the 'B' with the 'V', etc...

Having reached this stage on the segmental level, researchers are now working on improving the suprasegmentallevel (prosody); i.e., making the synthesized speech simulate emotions like anger, sorrow, etc... and be sounded like a question, or exclamation, or confirmation, etc... according to the semantics and pragmatics of speech.

The fundamental frequency is known to be an important element in prosody and is subject to researches aiming at analyzing its contour and linking it to the underlying speech. Results of these researches find their way to be applied to speech synthesizers in order to improve their prosody, i.e., making synthesized speech sound more natural.