摘要 |
PURPOSE: To make it easy to confirm the correction result of a parameter value by displaying an extracted phoneme parameter value on a screen, correcting the displayed parameter value with cursor keys or a mouse and obtaining a parameter interpolation value, and synthesizing a voice by using the parameter interpolation value. CONSTITUTION: A phoneme parameter value having a feature extracted is projected on the screen of a display device 10 such as a CRT. The projected parameter value is corrected with the cursor keys and mouse and the corrected parameter is stored as the interpolation value in a storage part 12. The stored interpolation value is processed by a voice synthesizing process part 6 to synthesize and output a voice. Thus, the interpolation value of the phoneme parameter can be viewed graphically, so the correction effect of the parameter value can easily be confirmed and precise correcting operation can be performed. |