Voice conversion method for a speech synthesis system comprises dividing a first speech time signal into temporary subsequent segments, folding the segments with a distortion time function and producing a second speech time signal

摘要

Voice conversion method comprises dividing a first speech time signal into temporary subsequent segments, folding the segments with a distortion time function and producing a second speech time signal by placing the folded segments together. Preferred Features: The first and/or second speech time signal are analogue or digital time signals. The first speech time signal is divided into quasi-periodic segments which are determined using a pitch tracking algorithm.