Improving model-based convolutive blind source separation techniques via bootstrap

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.

In this study, cross-talk cancellation in convolutive blind source separation (BSS) is investigated using a new post-processing algorithm. In this algorithm, exclusive activity periods (EAPs) are used to model cross-talks in which only one BSS output signal is assumed to be active. Inactive intervals of each EAP are used to estimate the cross-talk leaked from each active signal. Then, the estimated cross-talk is subtracted from BSS outputs. Simulation results show that the proposed algorithm successfully suppresses the crosstalk by improving the signal-to-interference ratio (SIR) and signal-to-distortion ratio (SDR) of BSS outputs about 12 and 5 dB, respectively.

We present a method of blind source separation (BSS) for speech signals using a complex extension of infinite sparse factor analysis (ISFA) in the frequency domain. Our method is robust against delayed signals that usually occur in real environments, such as reflections, short-time reverberations, and time lags of signals arriving at microphones. ISFA is a conventional non-parametric Bayesian method of BSS, which has only been applied to time domain signals because it can only deal with real signals. Our method uses complex normal distributions to estimate source signals and mixing matrix. Experimental results indicate that our method outperforms the conventional ISFA in the average signal-to-distortion ratio (SDR).

Spiny Lobster Sound Identification Based on Blind Source Separation (BSS) for Passive Acoustic Monitoring (PAM)

We propose a new algorithm for blind source separation (BSS), in which independent component analysis (ICA) and beamforming are combined to resolve the low-convergence problem through optimization in ICA. The proposed method consists of the following three parts: (1) frequency-domain ICA with direction-of-arrival (DOA) estimation; (2) null beamforming based on the estimated DOA; (3) integration of (1) and (2) based on the algorithm diversity in both iteration and frequency domain. The inverse of the mixing matrix obtained by ICA is temporally substituted by the matrix based on null beamforming through iterative optimization, and the temporal alternation between ICA and beamforming can realize fast- and high-convergence optimization. The results of the signal separation experiments reveal that the signal separation performance of the proposed algorithm is superior to that of the conventional ICA-based BSS method, even under reverberant conditions.

We propose a new algorithm for blind source separation (BSS), in which independent component analysis (ICA) and beamforming are combined to resolve the low-convergence problem through optimization in ICA. The proposed method consists of the following three parts: (1) frequency-domain ICA with direction-of-arrival (DOA) estimation, (2) null beamforming based on the estimated DOA, and (3) integration of (1) and (2) based on the algorithm diversity in both iteration and frequency domain. The inverse of the mixing matrix obtained by ICA is temporally substituted by the matrix based on null beamforming through iterative optimization, and the temporal alternation between ICA and beamforming can realize fast- and high-convergence optimization. The results of the signal separation experiments reveal that the signal separation performance of the proposed algorithm is superior to that of the conventional ICA-based BSS method, even under reverberant conditions.

Blind source separation (BSS) and sound activity detection (SAD) from a sound source mixture with minimum prior information are two major requirements for computational auditory scene analysis that recognizes auditory events in many environments. In daily environments, BSS suffers from many problems such as reverberation, a permutation problem in frequency-domain processing, and uncertainty about the number of sources in the observed mixture. While many conventional BSS methods resort to a cascaded combination of subprocesses, e.g., frequency-wise separation and permutation resolution, to overcome these problems, their outcomes may be affected by the worst subprocess. Our aim is to develop a unified framework to cope with these problems. Our method, called permutation-free infinite sparse factor analysis (PF-ISFA), is based on a nonparametric Bayesian framework that enables inference without a pre-determined number of sources. It solves BSS, SAD and the permutation problem at the same time. Our method has two key ideas: unified source activities for all the frequency bins and the activation probabilities of all the frequency bins of all the sources. Experiments were carried out to evaluate the separation performance and the SAD performance under four reverberant conditions. For separation performance in the BSS_EVAL criteria, our method outperformed conventional complex ISFA under all conditions. For SAD performance, our method outperformed the conventional method by 5.9–0.5% in F-measure under the condition RT20 = 30–600 [ms], respectively.

SUMMARY Maximum likelihood estimation of a vector regression parameter and variance com- ponents is considered for the mixed effects model when observations are right censored. A general scheme of estimation is given using the EM algorithm and detailed results found for the model with between and within block variation. This model is applied to the logarithms of survival times from a repeated measures design. In this paper we consider the estimation of the parameters when some of the observa- tions are right-censored. This can occur with repeated measurements on subjects where the response is a waiting time and the variability in log times is modelled by a normal distribution. Alternative models for this situation are given by Clayton & Cuzick (1985) and their methods involve partially parametric techniques using ranks and are based on models involving extreme value, log gamma and logistic distributions. We are concerned with use of the EM algorithm (Dempster, Laird & Rubin, 1977) to estimate the parameters of the mixed model. For uncensored data, Hartley & Rao (1967) have considered this solution. The analysis leads to a method which is straightforward to implement for the repeated measures model. Dempster et al. (1984) give an account of the method with uncensored observations. In ? 2 we consider a general implementation of the EM algorithm for the pure random effects model with censored data and then, in ? 3, we give explicit results for the simple random effects model. Section 4 uses these results for the analysis of repeated measures and ? 5 discusses some miscellaneous points arising. Finally, in ? 6, we give an example involving skin graft data, where there has been extensive analysis using matched pairs techniques.

This paper proposes a method for obtaining nonlinear models of distribution grid based on available measurements from the power grid. We formulate a maximum likelihood estimation (MLE) problem that estimates unknown line parameters—specifically, the impedance between nodes—using measured voltage magnitudes and net active and reactive power injections at each node. The nonlinear model for the distribution grid uses a nonlinear approximation of the DistFlow model, which includes line losses and is parameterized by the unknown line impedances. We solve the resulting MLE problem using an expectation maximization (EM) algorithm, tailored for the nonlinear model, and provide a numerically robust implementation. The proposed method is demonstrated on the IEEE 37-node test network, and we compare it with the state-of-the-art methods. The proposed method achieves a 70% reduction in voltage error and an error for state variables that is more than 10000 times smaller. A final comparison uses data from a real network, and the proposed method achieves parameter estimates with errors 100 times smaller than competing methods.

This paper introduces structural identification using expectation maximization (STRIDE), a novel application of the expectation maximization (EM) algorithm and approach for output-only modal identification. The EM algorithm can be used to estimate the maximum likelihood parameters of a state-space model. In this context, the state-space model represents the equation of motion for a linear dynamic system. STRIDE is an iterative procedure that uses Kalman filtering and Rauch-Tung-Striebel (RTS) smoothing equations to produce estimates of the unobserved states; these calculations are based on the observed data and prior estimates of the state-space parameters. With this information, the conditional likelihood of the model is maximized and the state-space parameters are updated at each iteration. Once an iteration meets user-prescribed convergence criterion, the algorithm ends—yielding maximum likelihood estimates (MLE) for the state-space model parameters. The modal properties of the structure are th...

The problem of blind source separation (BSS) for multiple-input multiple-output (MIMO) autoregressive (AR) mixtures is addressed in this paper. A new time-domain method for system identification and BSS is proposed based on the Gaussian mixture model (GMM) for sources distribution. The algorithm is based on the generalized expectation-maximization (GEM) method for joint estimation of the AR model parameters and the GMM parameters of the sources. The method is tested via simulations of synthetic and real audio signals. The results show that the proposed algorithm outperforms the well-known multidimensional linear predictive coding (LPC), and it achieves higher signal-to-interference ratio (SIR) in the BSS problem.

This paper deals with the problem of blind source separation in fMRI data analysis. Our main contribution is to present a maximum likelihood based method to blindly separate the brain activations in an fMRI experiment. Choosing the time frequency domain as the signal representation space, our method relies on the second order statistics and exploits the inter-source diversity. It is efficiently implemented by the EM (expectation-maximization) algorithm where the time courses of the brain activations are considered as the hidden variables. The estimation variance of the STFT (short time Fourier transform) is reduced by averaging across time frequency sub-domains. The successful separation of the right and left visual cortex activations during a visual fMRI experiment, in a block design, and the extraction of only the relevant tasks corroborate the effectiveness of our proposed separating algorithm.

Examining the robustness properties of maximum likelihood (ML) estimators of parameters in exponential power and generalized t distributions has been considered together. The well-known asymptotic properties of ML estimators of location, scale and added skewness parameters in these distributions are studied. The ML estimators for location, scale and scale variant (skewness) parameters are represented as an iterative reweighting algorithm (IRA) to compute the estimates of these parameters simultaneously. The artificial data are generated to examine performance of IRA for ML estimators of parameters simultaneously. We make a comparison between these two distributions to test the fitting performance on real data sets. The goodness of fit test and information criteria approve that robustness and fitting performance should be considered together as a key for modeling issue to have the best information from real data sets.

This paper deals with maximum-likelihood (ML) estimation of synchronization parameters for coded transmission systems. In particular, we present a unified framework based on both the sum-product (SP) algorithm and the expectation-maximization (EM) algorithm for the design of iterative synchronizers. The proposed approach is shown to encompass some known iterative synchronizers. In particular, we revisit a previously proposed framework based on the EM algorithm only by means of our "SP-EM" approach. The performance of the proposed synchronization method is assessed in terms of mean-square error and bit-error rate by simulation results. In particular, we consider the joint synchronization of the timing epoch and the carrier phase offset in the case of convolutionally coded and turbo-coded transmissions

In reliability analysis, an investigator is often interested in the assessment of a specific risk in presence of other risk factors. It is well known as the competing risks problem in the statistical literature. In this paper, we consider the analysis of incomplete data in the presence of competing risks among several groups. We mainly consider the latent failure times model formulation, and it is assumed that the lifetime distributions of the different latent failure times of a particular group follow Weibull distributions with different scale parameters, but the same shape parameter. Maximum likelihood estimators of the different parameters are obtained using a simple iterative procedure, and also by EM algorithm. Asymptotic distributions of the maximum likelihood estimators of the different parameters are obtained, and based on the asymptotic distributions, asymptotic confidence intervals are also proposed. Testing equality of the parameters among several groups is performed. One data set has been analysed for illustrative purposes.