Articulatory modeling: a possible role in concatenative text-to-speech synthesis

A brief introduction to articulatory modeling is presented. Methods of synthesis of speech with such models are described. Next, methods of controlling these synthesizers are considered, so as to produce continuous speech from text. Finally, the possibility of using articulatory models for concatenative text-to-speech is discussed. An important step in the realization of such synthesis is the acquisition of articulatory units. Some methods of deriving such units from natural speech are discussed.