摘要 |
PROBLEM TO BE SOLVED: To perform high-accuracy translation and to output a translated sentence in synthetic voice without feelings of incompatibility. SOLUTION: In a voice recognizing part 1, inputted voices are recognized, and the prosody information of inputted voices is extracted and supplied to a machine translation part 2 together with the voice recognized result of inputted voices. The machine translation part 2 has a translation table, in which language translation data for performing Japanese/English and English/Japanese translations are described together with prosody information concerning English and Japanese, and on the basis of that translation table, the voice recognized result is translated to a translated sentence with prosody information. This translated sentence with prosody information is supplied to a voice synthesizing part 3 and outputted in synthetic voices from a speaker 5 by performing ruled voice synthesizing processing. |