A METHOD FOR SPEECH SYNTHESIS HARDWARE, AND SPEECH SYNTHESIS APPARATUS

摘要

Speech synthesis is based on a human vocal chords model with two dynamically interlinked masses. Their positions determine glottis parameters. In particular, a first sub-model describes the dynamically interlinked masses acted upon by external forces. A second dynamical partial model describes dynamics of a contour of the glottis in dependence of geometry parameters produced by the first sub-model, and outputs a first part of the external forces in addition to transmitting the geometry parameters. A third hydrodynamic sub-model describes hydrodynamics in dependence of the transmitted parameters, is acted upon by air pressure parameters, and outputs a second part of the external forces and an actual volume velocity.