Sixth International Conference on Spoken Language Processing
(ICSLP 2000)

Beijing, China
October 16-20, 2000

Automatic Modeling and Implementation of Intonation for the Arabic Language in TTS Systems

Safa Nasser Eldin (1), Hanna Abdel Nour (2), Rajouani Abdenbi (3)

(1) Equipe Signal et Image, ENSERB Talence, France
(2) Al-Quds University, Electronic Engineering Department, Jerusalem, Palestine
(3) LESSA, Rabat, Morocco

This paper proposes a set of rules for the automatic generation of F0 contours for modern standard Arabic (MSA) affirmative and interrogative sentences. The objective is to finalize a model for the automatic processing of the intonative pattern in different TTS systems (e.g. synthesis by formants using Klatt synthesizer and synthesis by diphones, using a synthesizer based on PSOLA algorithm). The approach is based on the perceptive description of the Arabic intonation by extracting the pertinent variation of the intonative curve using a suitable stylization.

Full Paper

Bibliographic reference.  Eldin, Safa Nasser / Nour, Hanna Abdel / Abdenbi, Rajouani (2000): "Automatic modeling and implementation of intonation for the arabic language in TTS systems", In ICSLP-2000, vol.1, 597-600.