Sequential Unconstrained Minimization

(Hui Ling Lu)
(Google search: ``CCRMA Singing/Speech Synthesis'')

Klatt glottal (parabola) parameters are estimated jointly with vocal tract filter coefficients

• Formulation resembles that of the equation error method for system identification
• For phase alignement, we estimate
  • pitch (time varying)
  • glottal closure instant each period
• Optimization is convex in all but the phase-alignment dimension