Arabic Speech Synthesis Based on HMM
Krichi Mohamed Khalil, Adnan Cherif · 2018
This paper describes the synthesis of the Arab system based on Hidden Markov Models (HMM). HMMs have been successfully used in automatic speech recognition (ASR) from the 1970s, but recently they have been used for speech synthesis. Our synthesis system developed uses phonemes as HMM synthesis unit, Arabic database was developed for the first test. The main objective is to maintain the coherence of the consolidated text is interpreted by concatenating HMM phoneme. This method has several advantages. As it is parametric, it is possible to play on the HMM parameters change the characteristics of the voice producer. The developed model improves the quality of naturalness and intelligibility of speech synthesis in Arabic language environment.