Speech signals in frequency domain were separated based on discrete wavelet transform (DWT) and independent component analysis (ICA). First, mixed speech signals were decomposed into different frequency domains by DWT...Speech signals in frequency domain were separated based on discrete wavelet transform (DWT) and independent component analysis (ICA). First, mixed speech signals were decomposed into different frequency domains by DWT and the subbands of speech signals were separated using ICA in each wavelet domain; then, the permutation and scaling problems of frequency domain blind source separation (BSS) were solved by utilizing the correlation between adjacent bins in speech signals; at last, source signals were reconstructed from single branches. Experiments were carried out with 2 sources and 6 microphones using speech signals at sampling rate of 40 kHz. The microphones were aligned with 2 sources in front of them, on the left and right. The separation of one male and one female speeches lasted 2.5 s. It is proved that the new method is better than single ICA method and the signal to noise ratio is improved by 1 dB approximately.展开更多
A transformation formula containing four independent bases is found by a special inversion formula and it is applied to derive a few summation formulas for basic hypergeometric series only by elementary method. The hy...A transformation formula containing four independent bases is found by a special inversion formula and it is applied to derive a few summation formulas for basic hypergeometric series only by elementary method. The hypergeometric series, the limits of those formulas are also obtained.展开更多
The paper proposes a robust digital audio watermarking scheme using blind source separation(BSS) based on the global optimization of independency metric(IM),which is formulated as a generalized eigenvalue(GE) problem....The paper proposes a robust digital audio watermarking scheme using blind source separation(BSS) based on the global optimization of independency metric(IM),which is formulated as a generalized eigenvalue(GE) problem.Compared with traditional information-theoretical approaches used in digital audio watermarking,such as fast independent component analysis(FastICA),the proposed scheme has lower complexity without timeconsuming iteration steps used in FastICA.To make full use of the multiresolution characteristic of discrete wavelet transform(DWT) and the energy compression characteristic of discrete cosine transform(DCT),the watermark is embedded in the middle DWT-DCT coefficients and the independent component analysis(ICA) approach based on IM is used in the detecting scheme.Simulation results based on Stirmark for Audio v02 show that the proposed scheme has strong robustness as well as the imperceptibility and security.展开更多
基金Supported by Tianjin Municipal Science and Technology Commission (No.09JCYBJC02200)
文摘Speech signals in frequency domain were separated based on discrete wavelet transform (DWT) and independent component analysis (ICA). First, mixed speech signals were decomposed into different frequency domains by DWT and the subbands of speech signals were separated using ICA in each wavelet domain; then, the permutation and scaling problems of frequency domain blind source separation (BSS) were solved by utilizing the correlation between adjacent bins in speech signals; at last, source signals were reconstructed from single branches. Experiments were carried out with 2 sources and 6 microphones using speech signals at sampling rate of 40 kHz. The microphones were aligned with 2 sources in front of them, on the left and right. The separation of one male and one female speeches lasted 2.5 s. It is proved that the new method is better than single ICA method and the signal to noise ratio is improved by 1 dB approximately.
文摘A transformation formula containing four independent bases is found by a special inversion formula and it is applied to derive a few summation formulas for basic hypergeometric series only by elementary method. The hypergeometric series, the limits of those formulas are also obtained.
基金the Scientific Research Foundation for the Returned Overseas Chinese Scholars,State Education Ministry of Chinathe National Natural Science Foundation of China (No. 60802058)
文摘The paper proposes a robust digital audio watermarking scheme using blind source separation(BSS) based on the global optimization of independency metric(IM),which is formulated as a generalized eigenvalue(GE) problem.Compared with traditional information-theoretical approaches used in digital audio watermarking,such as fast independent component analysis(FastICA),the proposed scheme has lower complexity without timeconsuming iteration steps used in FastICA.To make full use of the multiresolution characteristic of discrete wavelet transform(DWT) and the energy compression characteristic of discrete cosine transform(DCT),the watermark is embedded in the middle DWT-DCT coefficients and the independent component analysis(ICA) approach based on IM is used in the detecting scheme.Simulation results based on Stirmark for Audio v02 show that the proposed scheme has strong robustness as well as the imperceptibility and security.
