History of speech synthesis

Text-to-speech has a long history: for example, the first autonomous automatic speech synthesis system in French, LIMSI's Icophone V, dates back to 1974. However, this is still a very active field of research. Current work focuses on both text comprehension and the reproduction of natural, personalized and expressive speech. Applications are multiplying. Readers wishing to explore the concepts presented in this article in greater depth can consult the references in the documentation section.

The general architecture of a synthesis system is thus made up of these two main parts (figure 1 ). The main modules corresponding to these processes are described in this article. Although all the examples quoted below are taken from French, it is important to emphasize...