摘要 |
PROBLEM TO BE SOLVED: To provide a voice signal compression device and a voice signal compression method capable of reducing a circuit scale or a processing load, and satisfactorily holding the sound quality of a decrypted voice signal with respect to a compression ratio.SOLUTION: A voice signal compression device 1 includes: a feature detection part 11 for detecting the feature of the signal component of a voice signal; and an error detection part 12 for detecting a difference between a plurality of features detected by the feature detection part 11 as an error. The error detection part 12 contrasts the sample of the feature as the prediction information of the signal component recorded on the basis of the already detected feature with a contrast feature as a feature to be remarked as a contrast object detected by the feature detection part 11, detects a difference between the sample and the contrast feature as an error, generates a new sample by adding the detected error to the sample, and uses the newly generated sample for its contrast with the contrast feature after generation. |