The goal for vocal tract modeling is to develop speech simulation of the human voice. To produce natural sounding simulated speech, the model of the vocal
tract must accurately represent the propagation of acoustic
waves and the proper characteristics of movement from one vocal tract shape
to another. The use of ultrasound, electron-beam computed tomography (EBCT)
and magnetic resonance imaging (MRI) have been used to measure the shape
of the vocal tract during sustained sound vocalization. The shape measurements
are important in generating speech simulation.
Applications