• Phonetics and Speech Sciences
• /
• v.7 no.4
• /
• pp.17-25
• /
• 2015

For blind source separation of convolutive mixtures, FDICA(Frequency Domain Independent Component Analysis) algorithms are generally used. Since FDICA algorithm such as Sawada FDICA, IVA(Independent Vector Analysis) works on the frequency bin basis with a natural gradient descent method, it takes much time to converge. In this paper, we propose a new method to improve convergence speed in FDICA algorithm. The proposed method reduces the number of iteration drastically in the process of natural gradient descent method by applying a weighted inner product constraint of unmixing matrix. Experimental results have shown that the proposed method achieved large improvement of convergence speed without degrading the separation performance of the baseline algorithms.

• Proceedings of the Korea Contents Association Conference
• /
• 2005.05a
• /
• pp.369-373
• /
• 2005

This paper proposes a variable step size parameter method in frequency domain ICA (FDICA). The FDICA and the temporal analysis (TA) algorithm are experimented for blind source separation (BSS). This paper will compare the separation qualities of these two algorithms in various reverberation environments. Furthermore, it is shown that the proposed technique has the better separation performance than those of two methods especially in recorded data.

• Proceedings of the Acoustical Society of Korea Conference
• /
• spring
• /
• pp.9-14
• /
• 2002

We propose a new algorithm for blind source separation (BSS), in which frequency-domain independent component analysis (FDICA) and time-domain ICA (TDICA) are combined to achieve a superior source-separation performance under reverberant conditions. Generally speaking, conventional TDICA fails to separate source signals under heavily reverberant conditions because of the low convergence in the iterative learning of the inverse of the mixing system. On the other hand, the separation performance of conventional FDICA also degrades significantly because the independence assumption of narrow-band signals collapses when the number of subbands increases. In the proposed method, the separated signals of FDICA are regarded as the input signals for TDICA, and we can remove the residual crosstalk components of FDICA by using TDICA. The experimental results obtained under the reverberant condition reveal that the separation performance of the proposed method is superior to that of conventional ICA-based BSS methods.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.8 no.10
• /
• pp.1451-1456
• /
• 2013

This paper propose a method for fast convergence technique in frequency domain independent component analysis (FDICA) using eigenanalysis. It important, such as SONAR system, to eliminate the interference sources through fast algorithm. Through eigenanalysing a two-by-two delayed mixture case, information of delay can be used for initial weighting parameters. Simulations show the improved performances in convergence speed and noise rejection rate. The proposed method can present close weights for optimal convergence, noise can be diminished drastically about 3 times epoch, and get the better resultss with 1~3dB than the conventional method.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.1
• /
• pp.14-20
• /
• 2006

In this paper. a novel technique to set up the initial weights in BSS of delayed mixtures is proposed. After analyzing Eigendecomposition for the correlation matrix of mixing data. the initial weights are set from the Eigenvectors ith delay information. The Proposed setting of initial weighting method for conventional FDICA technique improved the separation Performance. The computer simulation shows that the Proposed method achieves the improved SIR and faster convergence speed of learning curve.

• The Journal of the Korea Contents Association
• /
• v.6 no.12
• /
• pp.89-95
• /
• 2006

The performance of blind source separation(BSS) using independent component analysis (ICA) declines significantly in a reverberant environment. A post-processing method proposed in this paper was designed to remove the residual component precisely. The proposed method used modified NLMS(normalized least mean square) filter in frequency domain, to estimate cross-talk path that causes residual cross-talk components. Residual cross-talk components in one channel is correspond to direct components in another channel. Therefore, we can estimate cross-talk path using another channel input signals from adaptive filter. Step size is normalized by input signal power in conventional NLMS filter, but it is normalized by sum of input signal power and error signal power in modified NLMS filter. By using this method, we can prevent misadjustment of filter weights. The estimated residual cross-talk components are subtracted by non-stationary spectral subtraction. The computer simulation results using speech signals show that the proposed method improves the noise reduction ratio(NRR) by approximately 3dB on conventional FDICA.

• The Journal of the Acoustical Society of Korea
• /
• v.35 no.2
• /
• pp.110-117
• /
• 2016

Passive sonar systems detect and classify the target by analyzing the radiated noises from vessels. If multiple noise sources exist within the sonar detection range, it gets difficult to classify each noise source because mixture of noise sources are observed. To overcome this problem, a beamforming technique is used to separate noise sources spatially though it has various limitations. In this paper, we propose a new method that uses a FDICA (Frequency Domain Independent Component Analysis) to separate noise sources from the mixture. For experiments, each noise source signal was synthesized by considering the features such as machinery tonal components and propeller tonal components. And the results of before and after separation were compared by using LOFAR (Low Frequency Analysis and Recording), DEMON (Detection Envelope Modulation On Noise) analysis.

• Journal of Korea Multimedia Society
• /
• v.8 no.6
• /
• pp.753-760
• /
• 2005

This paper introduces a variable step-size adaptive algorithm for blind source separation. From the frequency characteristics of mixed input signals, we need to adjust the convergence speed regularly in each frequency bin. This algorithm varies a step-size according to the magnitude of input at each frequency bin. This guarantee of the regular convergence in each frequency bin would become more efficient in separation performances than conventional fixed step-size FDICA. Computer simulation results show the improvement of about 5 dB in signal to interference ratio (SIR) and the better separation quality.

• The Journal of the Acoustical Society of Korea
• /
• v.25 no.2E
• /
• pp.36-42
• /
• 2006

In this paper, a novel post processing using Wiener filtering technique is proposed to p rm further interference reduction in FDICA. Using the proposed method, the target signal components are remained with little attenuation while the interference components are drastically suppressed. The results of experiments show that the proposed method achieves a reduction of the residual crosstalk. Compared to the NLMS method, the proposed method has slightly better separation performance in SIR, and even requires much less computational complexity.