摘要 |
<p>A speech processing device in which prediction taps for finding prediction values of a speech of high sound quality are extracted from a synthesized sound obtained based on linear prediction coefficients and residual signals, generated from a preset code. The device includes a prediction tap extracling unit (45) for extracting, from the synthesized sound and from said code or the information derived from said code, the prediction taps used for predicting the speech of high sound quality, as target speech, and a dass tap extraction unit (46) for extracting class taps, used for classifying the target speech to one of a plurality of classes, from the above code and said synthesized sound. The device has an acquisition unit for acquiring the tap coefficients associated with the dass of the target speech from among the tap coefficients as found by previous learning.</p> |