摘要 |
PROBLEM TO BE SOLVED: To increase the performance of speech recognition for speech data related to an image while taking microphone characteristics, an ambient sound, etc., into consideration. SOLUTION: A sound model corresponding to information which is included in an inputted image file and specifies photographing equipment is selected out of a plurality of prestored sound models (S503, S504) with reference to a table in which information specifying photographing equipment and information specifying a sound model corresponding thereto are described (S502), and based upon the selected sound model, speech recognition processing is carried out for speech data included in the inputted image file. COPYRIGHT: (C)2005,JPO&NCIPI
|