A new F0 contour control method based on vector representation of F0 contour

This paper proposes a new fundamental frequency(F0) contour control method based on vector representation of F0 contour. The main points of the proposed method are as follows; (1) Desired F0 contours are created by selecting or modifying natural F0 contours held in a speech database. (2) F0 contour selection is based on statistical estimation using a vector representation of F0 contour (3) The selected F0 contour is modified to match the target context according to rules produced by statistical learning. An evaluation by listening tests confirms the superior performance of our proposed over the conventional method approach to F0 modeling.

[1]  D. R. Ladd,et al.  Manipulating synthetic intonation for speaker characterisation , 1991, [Proceedings] ICASSP 91: 1991 International Conference on Acoustics, Speech, and Signal Processing.

[2]  Thierry Dutoit,et al.  Automatic prosody generation using suprasegmental unit selection , 1998, SSW.

[3]  H. Sato,et al.  Two-stage F/sub 0/ control model using syllable based F/sub 0/ units , 1992, [Proceedings] ICASSP-92: 1992 IEEE International Conference on Acoustics, Speech, and Signal Processing.

[4]  Chikio Hayashi On the quantification of qualitative data from the mathematico-statistical point of view , 1950 .

[5]  Yoshinori Sagisaka,et al.  Fundamental frequency database with linguistic and phonetic information , 1989 .

[6]  Nick Campbell Prosody and the selection of units for concatenation synthesis , 1994, SSW.