To provide a technique which can transmit a feeling while protecting the privacy of a person who speaks words in time to the input of voice data without displaying the face of the person who voices the words.
A speech recognition part 11 extracts a featured frequency component of voice data inputted through a microphone M1 and a speech classification part 12 classifies the data into preset reference speech data including language sound data and feeling data. Then an image data selection part 13 selects character expression information data corresponding to the reference voice data from a storage part 14 and displays an image on a display screen D1. A voice processing part 16, on the other hand, converts the voice data inputted from the microphone M1 to generate a voice through a speaker S1. At this time, the image display and voice output are synchronized with each other under the control of a control part 20. The image display and voice output can be carried out even on game terminal devices G2 to G4 at communication destinations as well.
YASUNAKA ATSUSHI
SUGIMOTO TAKASHI
HASEGAWA HIROSHI
