VOICE SIGNAL PROCESSING DEVICE AND PROGRAM

摘要

PROBLEM TO BE SOLVED: To adjust the magnitude of a background sound signal so as to get easily listenable balance of voice and background sound by accurately detecting a voice section and converting speech speed with good sound quality.SOLUTION: A majority vote determination unit 5 of a voice signal processing device 10 performs majority vote determination for voice consecutive section information P1, P2 and subtitle display section information P3 according to prescribed weighting, and generates correction voice section information CP. When the correction voice section information CP indicates a voice section, a speech speed conversion unit 8 converts speech speed to prescribed speed by extending or compacting an input voice signal NplsBG and an estimated voice signal N', and the like in units of one fundamental cycle f extracted from the estimated voice signal N', and when the correction voice section information CP indicates a non-voice section, the speech speed conversion unit 8 does not perform speed conversion or converts the speech speed to the prescribed speed. A mixture ratio adjusting unit 9 controls the magnitude of background sound, for example, by a background sound suppression technique in the voice section and by a gain control technique in the non-voice section regarding the voice section and the non-voice section indicated by the correction voice section information CP.