Voice conversion by prosody and vocal tract modification

In this paper we proposed some exible methods, which are useful in the process of voice conversion. The pro-posed methods modify the shape of the vocal tract system and the characteristics of the prosody according to the de-sired requirement. The shape of the vocal tract system is modied by shifting the major resonant frequencies (for-mants) of the short term spectrum, and altering their band-widths accordingly. In the case of prosody modication, the required durational and intonational characteristics are im-posed on the given speech signal. In the proposed method, the prosodic characteristics are manipulated using instants of signicant excitation. The instants of signicant excita-tion correspond to the instants of glottal closure (epochs) in the case of voiced speech, and to some random excita-tions like onset of burst in the case of nonvoiced speech. Instants of signicant excitation are computed from the Lin-ear Prediction (LP) residual of the speech signals by using the property of average group delay of minimum phase sig-nals. The manipulations of durational characteristics and pitch contour (intonation pattern) are achieved by manipu-lating the LP residual with the help of the knowledge of the instants of signicant excitation. The modied LP residual is used to excite the time varying lter. The lter parameters are updated according to the desired vocal tract characteris-tics. The proposed methods are evaluated using listening tests. 1