Abstract: A method for separating a blind signal includes: converting mixed signals of a time domain collected by using a plurality of sensors into mixed signals of a frequency domain; calculating a separation filter from the mixed signals which have been converted into those of the frequency domain; calculating an inverse filter of the separation filter; calculating the difference in phase between the respective sensors from the calculated inverse filter; permutation-sorting the separation filter by using the calculated phase difference; and separating the mixed signals of the frequency domain by using the permutation-sorted separation filter.

Abstract: A method for separating a blind signal includes: converting mixed signals of a time domain collected by using a plurality of sensors into mixed signals of a frequency domain; calculating a separation filter from the mixed signals which have been converted into those of the frequency domain; calculating an inverse filter of the separation filter; calculating the difference in phase between the respective sensors from the calculated inverse filter; permutation-sorting the separation filter by using the calculated phase difference; and separating the mixed signals of the frequency domain by using the permutation-sorted separation filter.

Abstract: A method and apparatus performing blind source separation using frequency-domain normalized multichannel blind deconvolution. Multichannel mixed signals are frames of N samples including r consecutive blocks of M samples. The frames are separated using separating filters in frequency domain in an overlap-save manner by discrete Fourier transform (DFT). The separated signals are then converted back into time domain using inverse DFT applied to a nonlinear function. Cross-power spectra between separated signals and nonlinear-transformed signals are computed and normalized by power spectra of both separated signals and nonlinear-transformed signals to have flat spectra. Time domain constraint is then applied to preserve first L cross-correlations. These alias-free normalized cross-power spectra are further constrained by nonholonomic constraints. Then, natural gradient is computed by convolving alias-free normalized cross-power spectra with separating filters.

Abstract: A method and apparatus performing blind source separation using frequency-domain normalized multichannel blind deconvolution. Multichannel mixed signals are frames of N samples including r consecutive blocks of M samples. The frames are separated using separating filters in frequency domain in an overlap-save manner by discrete Fourier transform (DFT). The separated signals are then converted back into time domain using inverse DFT applied to a nonlinear function. Cross-power spectra between separated signals and nonlinear-transformed signals are computed and normalized by power spectra of both separated signals and nonlinear-transformed signals to have flat spectra. Time domain constraint is then applied to preserve first L cross-correlations. These alias-free normalized cross-power spectra are further constrained by nonholonomic constraints. Then, natural gradient is computed by convolving alias-free normalized cross-power spectra with separating filters.