| 摘要 |
PROBLEM TO BE SOLVED: To easily improve the SN ratio and the sound quality in feeling of hearing by converting a spectrum, output signals of a spectrum addition means, into time-domain signals. SOLUTION: Voice/unvoice judgment part 32 judges whether a frame-split speech signal is voiced or unvoiced, adding the judged information to the speech signal. A Fourier transform part 33 performs Fourier transform (FFT), analyzing spectral components and outputting spectral information. A noise spectrum estimation part 35 estimates noise spectrum included in the output by referring to a spectrum memory 36. An amplitude spectrum subtraction part 34 subtracts from the out put of Fourier transformation part 33 what noise spectrum estimated from the noise spectrum estimation part 35 is multiplied by a subtraction coefficient. And, an inverse-Fourier transformation part 40, which is a timedomain signal transformation means processes the output of the amplitude spectrum subtraction part 34 with inverseFourier transform (IFFT). |
