| 摘要 |
<p><P>PROBLEM TO BE SOLVED: To provide a speech authenticating apparatus which increases a matching rate. <P>SOLUTION: A speech collating apparatus 10 is equipped with a plurality of speech input means for inputting the speech signal of a collated speaker, a speech comparing means of comparing a plurality of speech signals inputted by each of the plurality of speech input means with one another, a speech extracting means of extracting at least two approximate speech signals from the plurality of speech signals according to the comparison result of the speech comparing means, a speech correcting means of correcting at least one of the two approximate speech signals according to at least the two approximate speech signals extracted by the speech extracting means, and a speech recognizing means of analyzing a voiceprint and a phoneme according to the speech signal corrected by the speech correcting means, and has a function of collating the speech signal of the collated speaker with a speech signal of a registered speaker according to the results of the voiceprint and phoneme analyses carried out by the speech recognizing means. <P>COPYRIGHT: (C)2004,JPO</p> |
