Abstract
As speech synthesis systems are now widely used for many apparatus and services. In recent years, speech synthesis systems are demanded to have various vocal quality and tones according to a user's request or a use. Formant synthesis can change a tone of voice very freely by adjusting parameter values of voice feature. This paper proposes to use interactive evolutionary computing to adjust those parameters. In conclusion, the present study has demonstrated that formant synthesis using genetic algorithm can generate synthetic voices with various tones but it does not fully come to satisfy user's demand.
