Kitakyushu Foundation for the Advancement of Industry, Science and Technology

Hiromu Gotanda

Keiichi Kaneda

Takeshi Koya

The present invention provides a method for recovering target speech based on shapes of amplitude distributions of split spectra obtained by use of blind signal separation. This method includes: a first step of receiving target speech emitted from a sound source and a noise emitted from another sound source and forming mixed signals of the target speech and the noise at a first microphone and at a second microphone; a second step of performing the Fourier transform of the mixed signals from the time domain to the frequency domain, decomposing the mixed signals into two separated signals Uand Uby use of the Independent Component Analysis, and, based on transmission path characteristics of the four different paths from the two sound sources to the first and second microphones, generating the split spectra v, v, vand vfrom the separated signals Uand U; and a third step of extracting estimated spectra Z* corresponding to the target speech to generate a recovered spectrum group of the target speech, wherein the split spectra v, v, v, and vare analyzed by applying criteria based on the shape of the amplitude distribution of each of the split spectra v, v, v, and v, and performing the inverse Fourier transform of the recovered spectrum group from the frequency domain to the time domain to recover the target speech.

Method for recovering target speech based on amplitude distributions of separated signals

Method for recovering target speech by extracting signal components falling in a speech segment, which is determined based on separated signals obtained through the Independent Component Analysis, thereby minimizing the residual noise in the recovered target speech. The present method comprises: the first step of receiving target speech emitted from a sound source and a noise emitted from another sound source and extracting estimated spectra Y* corresponding to the target speech by use of the Independent Component Analysis; the second step of separating from the estimated spectra Y* an estimated spectrum series group y* in which the noise is removed by applying separation judgment criteria based on the kurtosis of the amplitude distribution of each of estimated spectrum series in Y*; the third step of detecting a speech segment and a noise segment of the total sum F of all the estimated spectrum series in y* by applying detection judgment criteria based on a predetermined threshold value T that is determined by the maximum value of F; and the fourth step of extracting components falling in the speech segment from the estimated spectra Y* to generate a recovered spectrum group of the target speech for recovering the target speech.

Growing community of inventors

Fukuoka, Japan

Hiromu Gotanda