| 摘要 |
<P>PROBLEM TO BE SOLVED: To generate speech synthesis information that is capable of synthesizing natural sounds. <P>SOLUTION: A storage device 12 stores speech synthesis information S that includes phoneme string information SA and feature quantity information SB. The phoneme string information SA assigns a sound length a3 to each phoneme of sound to be synthesized. The feature quantity information SB assigns a time change of a feature pitch of the sound to be synthesized. A display control part 22 causes a display device 16 to display a phoneme string image 32 showing a time series of phonemes and a feature quantity image 34 showing the time change of the feature quantity. An editing processing part 24 edits the sound length a3 of each phoneme σ[n] assigned by the phoneme string information SA with a degree of expansion and contraction K[n] according to a pitch P[n] of the phoneme σ[n] assigned by the feature quantity information SB. <P>COPYRIGHT: (C)2012,JPO&INPIT |
