PURPOSE: A voice recognition system and method are provided to recognize the speech word of a user with the voice and vision information of the user. CONSTITUTION: A voice-based voice recognition unit(110) receives the user voice from an input module. The voice-based voice recognition unit creates recognition word information. An image-based voice recognition unit(130) decides the selection of the recognition word information on the basis of the recognition word information and the vision information. In case the recognition word information is selected, the image-based voice recognition unit decides a recognized word as an effective word. The image-based voice recognition unit offers effective word information to an external device.
申请公布号
KR20110069605(A)
申请公布日期
2011.06.23
申请号
KR20090126406
申请日期
2009.12.17
申请人
KOREA INSTITUTE OF INDUSTRIAL TECHNOLOGY
发明人
KO, WOO HYUN;JI, SANG HOON;NAM, KYUNG TAE;LEE, SANG MOO;SHON, WOONG HEE