摘要 |
PROBLEM TO BE SOLVED: To prevent an increase in the execution load of voice processing in a voice processing system. ! SOLUTION: A voice processing device 51-1 comprises a voice input unit 511, a speaker position identification unit 513, a group creation unit 514, a signal processing unit 515, and a voice recognition unit 516. The voice input unit 511 respectively receives inputs of voice signals that are based on voice of respective speakers. The speaker position identification unit 513 identifies the positions of respective speakers. The group creation unit 514 selects, from among a plurality of speakers, a plurality of speakers positioned in a predetermined range as one group. The signal processing unit 515 performs signal processing for resolving voice crosstalk on voice signals of the plurality of speakers in the group. The voice recognition unit 516 performs voice recognition processing on the voice signals on the basis of a result of the signal processing. ! COPYRIGHT: (C)2015,JPO&INPIT |