Using instantaneous frequency and aperiodicity detection to estimate FO for high-quality speech synthesis

   Abstract