摘要 |
<p><P>PROBLEM TO BE SOLVED: To provide a voice recognition device and a voice recognition method capable of preventing misunderstanding of an alphabet. <P>SOLUTION: When an operator voice-inputs a specified number of a part, the segments of an inputted voice is voice-recognized. When a control part judges that the inputted voice of "roku" does not correspond to identification information (S302:N), an identification result of "6" being a number is given (S305), for example. When the control part judges that the inputted voice of "arufa" corresponds to identification information (S302:Y), it extracts an alphabet "A" corresponding to identification information from a data storage part 201 (S303) and recognizes it as an English character "A". <P>COPYRIGHT: (C)2004,JPO</p> |