Adding emotional factors to synthesized voices

摘要

This paper deals with how to add emotional factors to synthesized voices by applying essentic forms of emotion which reflect the timewise response of muscle reaction to each kind of emotions. We modulate the magnitude and fundamental frequency of given series of voice with no emotional factor by the essentic standard forms obtained by fingertip pressure. Psychological evaluation of these modulated voices shows that anger, sadness, hate and joy are well regenerated by those essentic forms with appropriate multiplication factor but love, reverence and sex are not succeeded to be expressed by this method.