| 摘要 |
<p>PURPOSE:To improve recognition accuracy and processing time by restricting the dictionary categories matched with the command hierarchy of an application using two identifiers(IDs). CONSTITUTION:Each voice dictionary has two IDs, the 1st ID1 indicates a self-dictionary category to which the word dictionary belongs and the 2nd ID2 indicates the succeeding dictionary category when the word concerned is recognized. Since the recognition operation of all words is executed and only the word dictionary having the same ID1 as the ID2 of the recognized word is used for the succeeding recognition processing in an conventional method, the number of words to be recognized is restricted. Thus, the number of dictionaries to be recognized for a voice key emulator can be controlled so as to be always minimized and the recognition accuracy and the processing time can be improved.</p> |
