A pitch-contour generation method combining ANN, global variance, and real-contour selection

Pitch contours are important for synthesizing highly natural speech signal. In this paper, we study a new pitch-contour generation method. The proposed method combines ANN prediction module with global-variance matching (GVM) and real contour selection (RCS) modules. A syllable pitch contour is first analyzed and then transformed to a DCT-coefficient vector via discrete cosine transform (DCT). Each sequence of DCT vectors analyzed from a training sentence plus contextual parameters is then used to train the ANN weights and GVM parameters. In pitch-contour generation experiments, we measure variance-ratio (VR) values for objective evaluations. The modules, i.e. GVM and RCS, are shown to be helpful to promote VR values. In addition, in subjective evaluation, the pitch-contour generation method, i.e. ANN + GVM, is shown to be more natural than the method only using ANN. Moreover, the ANN + GVM + RCS method is shown to be better than ANN + GVTVL.