Research idea

Research idea Could you use a robotics approach for text-to-speech? Build a parametrized model of bones and muscles, and teach it to speak. Or maybe this approach is subsumed in a neural net approach: the parameters are free to model the mouth if that’s an optimal model…