An Arabic Text-To-Speech System Based on Artificial Neural Networks

Problem statement: With the rapid advancement in information technology and communications, computer systems increasingly offer the users the opportunity to interact with information through speech. The interest in speech synthesis and in building voices is increasing. Worldwide, speech synthesizers have been developed for many popular languages English, Spanish and French and many researches and developments have been applied to those languages. Arabic on the other hand, has been given little attention compared to other languages of similar importance and the research in Arabic is still in its infancy. Based on these ideas, we introduced a system to transform Arabic text that was retrieved from a search engine into spoken words. Approach: We designed a text-to-speech system in which we used concatenative speech synthesis approach to synthesize Arabic text. The synthesizer was based on artificial neural networks, specifically the unsupervised learning paradigm. Different sizes of speech units had been used to produce spoken utterances, which are words, diphones and triphones. We also built a dictionary of 500 common words of Arabic. The smaller speech units (diphones and triphones) used for synthesis were chosen to achieve unlimited vocabulary of speech, while the word units were used for synthesizing limited set of sentences. Results: The system showed very high accuracy in synthesizing the Arabic text and the output speech was highly intelligible. For the word and diphone unit experiments, we could reach an accuracy of 99% while for the triphone units we reached an accuracy of 86.5%. Conclusion: An Arabic text-to-speech synthesizer was built with the ability to produce unlimited number of words with high quality voice.