Joint Independent Subspace Analysis: A Quasi-Newton Algorithm

Independent vector analysis (IVA) is a special form of independent component analysis (ICA), which has demonstrated its prominent performance in solving convolutive blind source separation (BSS) problems in the frequency domain. Most IVA algorithms are based on optimizing certain contrast functions, where the main difficulty of these approaches lies in finding a reliable and fast estimation of the unknown distribution of sources. Despite the rich availability of efficient tensorial approaches to the standard ICA problem, these methods have not been explored considerably for IVA. In this article, we propose a matrix joint diagonalization approach to solve the complex IVA problem. The new factorization neither relies on a whitening process, nor does it require an estimate of the joint probability distribution of the dependent signal groups. The latter is in contrast to most IVA approaches up to date. The underlying geometry of the problem is investigated together with a critical point analysis of the resulting cost function. A conjugate gradient algorithm on the appropriate manifold setting is developed.

A new independent component analysis (ICA) formulation called independent vector analysis (IVA) was proposed in order to solve the permutation problem in convolutive blind source separation (BSS). Instead of running ICA in each frequency bin separately and correcting the disorder with an additional algorithmic scheme afterwards, IVA exploited the dependency among the frequency components of a source and dealt with them as a multivariate source by modeling it with sparse and spherically, or radially, symmetric joint probability density functions (pdfs). In this paper, we compare the speech separation performances of IVA by using a group of l <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">p</sup> -norm-invariant sparse pdfs where the value of and the sparseness can be controlled. Also, we derive an IVA algorithm from a nonparametric perspective with the constraint of spherical symmetry and high dimensionality. Simulation results confirm the efficiency of assuming sparseness and spherical symmetry for the speech model in the frequency domain.

Fast fixed-point independent vector analysis algorithms for convolutive blind source separation

The independent vector analysis (IVA) algorithm employs a multivariate source prior to retain the dependency between different frequency bins of each source and thereby avoids the permutation problem that is inherent to blind source separation (BSS). In this paper, a multivariate Student's t distribution is adopted as the source prior, which because of its heavy tail nature can better model the large amplitude information in the frequency bins. Therefore it can improve the separation performance and the convergence speed of the IVA and fast version of the IVA (FastIVA) algorithms as compared with the IVA algorithm based on another multivariate super Gaussian source prior. Separation performance with real binaural room impulse responses (BRIRs) is evaluated by detailed simulation studies when using the different source priors, and the experimental results confirm that the IVA and the FastIVA with the proposed multivariate Student's t source prior can consistently achieve improved and faster separation performance.

We present a new approach to independent component analysis (ICA) by extending the formulation of univariate source signals to multivariate source signals. The new approach is termed independent vector analysis (IVA). In the model, we assume that linear mixing model exists in each dimension separately, and the latent sources are independent of the others. In contrast to ICA, the sources are random vectors, not just single variables, which means the elements of a random vector are closely related to the others. Thus, we assume the dependency between the elements of a source vector. In this manner, we define dependence between vectors as Kullback-Leibler divergence between the total joint probability of vectors and the product of marginal probabilities of vectors. Then, the model allows independence between multivariate source signals represented as random vectors, and dependence between the source signals within the vector representation. The proposed vector density model can for example capture variance dependencies within a vector source signal. There are several applications of this new formulation. In the separation of acoustic sources, the algorithm mitigates the permutation problem, i.e. the usual ICA algorithms applied in to the frequency domain mixture data suffer from the unknown permutation of the output signals. Although there are several engineering solutions to fix this problem after the ICA stage, the proposed method provides a natural solution to the problem by capturing the inherent dependencies of the acoustic signals. It therefore avoids the permutation problem and allows the separation of sources in very challenging environments for many sound sources.

Complex-valued independent vector analysis: Application to multivariate Gaussian model

A novel method is proposed to improve the performance of independent vector analysis (IVA) for blind signal separation of acoustic mixtures. IVA is a frequency-domain approach that successfully resolves the well-known permutation problem by applying a spherical dependency model to all pairs of frequency bins. The dependency model of IVA is equivalent to a single clique in an undirected graph; a clique in graph theory is defined as a subset of vertices in which any pair of vertices is connected by an undirected edge. Therefore, IVA imposes the same amount of statistical dependency on every pair of frequency bins, which may not match the characteristics of real-world signals. The proposed method allows variable amounts of statistical dependencies according to the correlation coefficients observed in real acoustic signals and, hence, enables more accurate modeling of statistical dependencies. A number of cliques constitutes the new dependency graph so that neighboring frequency bins are assigned to the same clique, while distant bins are assigned to different cliques. The permutation ambiguity is resolved by overlapped frequency bins between neighboring cliques. For speech signals, we observed especially strong correlations across neighboring frequency bins and a decrease in these correlations with an increase in the distance between bins. The clique sizes are either fixed, or determined by the reciprocal of the mel-frequency scale to impose a wider dependency on low-frequency components. Experimental results showed improved performances over conventional IVA. The signal-to-interference ratio improved from 15.5 to 18.8 dB on average for seven different source locations. When we varied the clique sizes according to the observed correlations, the stability of the proposed method increased with a large number of cliques.

Independent vector analysis followed by HMM-based feature enhancement for robust speech recognition

Independent vector analysis (IVA) is an attractive solution to address the problem of joint blind source separation (JBSS), that is, the simultaneous extraction of latent sources from several datasets implicitly sharing some information. Among IVA approaches, we focus here on the celebrated IVA-G model, that describes observed data through the mixing of independent Gaussian source vectors across the datasets. IVA-G algorithms usually seek the values of demixing matrices that maximize the joint likelihood of the datasets, estimating the sources using these demixing matrices. Instead, we write the likelihood of the data with respect to both the demixing matrices and the precision matrices of the source estimates. This allows us to formulate a cost function whose mathematical properties enable the use of a proximal alternating algorithm based on closed form operators with provable convergence to a critical point. After establishing the convergence properties of the new algorithm, we illustrate its desirable performance in separating sources with covariance structures that represent varying degrees of difficulty for JBSS.

The issue of convolutive blind source separation (BSS) is addressed in this paper. Independent low-rank matrix analysis (ILRMA), unifying frequency-domain independent component analysis (FDICA) and nonnegative matrix factorization (NMF), is a method that has recently proposed to model low-rank structure of source spectra by using NMF in addition to independence between sources used in FDICA and independent vector analysis (IVA). Although ILRMA has been shown to provide better separation performance than FDICA and IVA, the frequency components of each source are assumed to be independent in ILRMA due to NMF modeling of source spectra, which may degrade its performance when the short-term Fourier transform (STFT) is unable to decorrelate the frequency components for each source. This paper therefore presents a new BSS method that unifies IVA and positive semidefinite tensor factorization (PSDTF). PSDTF models not only power spectra in the same way NMF does but also models the correlations between frequency bins in each source. The proposed method can be viewed as a multichannel extension of PSDTF and exploits both the independence between sources and the inter-frequency correlations as a clue for separating mixtures. Experimental results indicate the improved performance of our approach.

Electrocardiogram (ECG) signals are normally contaminated by various physiological and nonphysiological artifacts. Among these artifacts baseline wandering, electrode movement and muscle artifacts are particularly difficult to remove. Independent component analysis (ICA) is a well-known technique of blind source separation (BSS) and is extensively used in literature for ECG artifact elimination. In this article, the independent vector analysis (IVA) is used for artifact removal in the ECG data. This technique takes advantage of both the canonical correlation analysis (CCA) and the ICA due to the utilization of second-order and high order statistics for un-mixing of the recorded mixed data. The utilization of recorded signals along with their delayed versions makes the IVA-based technique more practical. The proposed technique is evaluated on real and simulated ECG signals and it shows that the proposed technique outperforms the CCA and ICA because it removes the artifacts while altering the ECG signals minimally.

We propose a blind source separation method that yields high-quality speech with low distortion. Time-frequency (TF) masking can effectively reduce interference, but it produces nonlinear distortion. By contrast, linear filtering using a separation matrix such as independent vector analysis (IVA) can avoid nonlinear distortion, but the separation performance is reduced under reverberant conditions. The tandem connectionist approach combines several separation methods and it has been used frequently to compensate for the disadvantages of these methods. In this study, we propose associative memory model (AMM)-based linear filtering and a tandem connectionist framework, which applies TF masking followed by linear filtering. By using AMM trained with speech spectra to optimize the separation matrix, the proposed linear filtering method considers the properties of speech that are not considered explicitly in IVA, such as the harmonic components of spectra. TF masking is applied in the proposed tandem connectionist framework to reduce unwanted components that hinder the optimization of the separation matrix, and it is approximated by using a linear separation matrix to reduce nonlinear distortion. The results obtained in simultaneous speech separation experiments demonstrate that although the proposed linear filtering method can increase the signal-to-distortion ratio (SDR) and signal-to-interference ratio (SIR) compared with IVA, the proposed tandem connectionist framework can obtain greater increases in SDR and SIR, and it reduces the phoneme error rate more than the proposed linear filtering method.

An introduction of multi-user multiple-input multiple-output (MU-MIMO) to distributed antenna-based small-cell networks is considered a promising approach toward 5G mobile networks to enhance the sum throughput. However, the inter-symbol interference (ISI) caused by the channel frequency-selectivity, the inter-antenna interference (IAI), and the inter-user interference (IUI) degrade the MU-MIMO downlink throughput. This paper proposes two joint transmit-and-receive (Tx/Rx) filtering schemes (called BD-SVD and MMSE-SVD) for MU-MIMO downlinks using single-carrier (SC) transmission and orthogonal frequency-division multiplexing (OFDM) transmission. In BD-SVD, IUI is removed by block diagonalization (BD) at the transmitter side, and then, the equivalent channel after BD is transformed into IAI-free eigenmodes using singular value decomposition (SVD). On the other hand, in MMSE-SVD, BD is not used and the channel is directly transformed into eigenmodes. IUI and IAI are suppressed by a minimum mean square error (MMSE)-based precoding at the transmitter assuming that each receiver does eigenmode reception. Furthermore, in the case of SC downlink, ISI is suppressed by applying MMSE-based Tx power allocation (PA) and Rx frequency-domain equalization (FDE) to each eigenmode. Numerical results show that BD-SVD and MMSE-SVD achieve a higher sum throughput than conventional MMSE precoding.

A family of new MMSE blind channel equalization algorithms based on second order statistics are proposed. Instead of estimating the channel impulse response, we directly estimate the cross-correlation function needed in Wiener-Hopf filters. We develop several different schemes to estimate the cross-correlation vector, with which different Wiener filters are derived according to minimum mean square error (MMSE). Unlike many known subspace methods, these equalization algorithms do not rely on signal and noise subspace separation and are consequently more robust to channel order estimation errors. Their implementation requires no adjustment for either single or multiple user systems. They can effectively equalize single-input multiple-output (SIMO) systems and can reduce the multiple-input multiple-output (MIMO) systems into a memoryless signal mixing system for source separation. The implementations of these algorithms on SIMO system are given and simulation examples are provided to demonstrate their superior performance over some existing algorithms.

A new method for solving the permutation problem in the frequency-domain Blind Source Separation(BSS) was presented.This new method divided the whole frequency-domain into three sections and applied different permutation algorithms to different sections.The neighboring frequency correlation and harmonic frequency correlation coefficient of signal amplitudes were used together for the low-frequency section,the direction of arrival estimation for speech sources for the mid-frequency section,and the combination of correlation coefficient comparison and direction estimation for the high-frequency section,respectively.Experimental results show that the new method provides a more robust and precise solution to the permutation problem than the algorithm only with interfrequency correlation coefficient of signal amplitudes or only with direction of arrival estimation for sources.