| 摘要 |
<p>A vocoder based voice recognizer recognizes a spoken word using linear prediction coding based vocoder data without completely reconstructing the voice data. The recognizer generates at least one energy estimate per frame of the vocoder data (60) and searches for word boundaries in the vocoder data (64) using the associated energy estimates. If a word is found (66), the linear prediction coding word parameters are extracted (68) from the vocoder data associated with the word and recognition features are calculated (70) from the extracted linear prediction coding word parameters. Finally, the recognition features are matched with previously stored recognition features of other words (40), thereby recognizing the spoken word.</p> |
