Building diphone database for Arabic text to speech synthesis system

The need to integrate automatic communication systems in daily applications and technology progress has encouraged speech synthesis sector development. Many variety of voice systems synthesis exist but diphones concatenation method appears to be efficient to contribute to speech synthesis of many languages. The first task in any development of speech synthesis system is to prepare the data base used for this aim. This paper describes the process of building an Arabic speech corpus based on diphones. We developed an acoustic data base of arabic diphones by studying the different possible combinations of phonemes to list all the possible diphones.