Abstract:Maximum Likelihood (ML) estimation requires precise knowledge of the underlying statistical model. In Quasi ML (QML), a presumed model is used as a substitute to the (unknown) true model. In the context of Independent Vector Analysis (IVA), we consider the Gaussian QML Estimate (QMLE) of the demixing matrices set and present an (approximate) analysis of its asymptotic separation performance. In Gaussian QML the sources are presumed to be Gaussian, with covariance matrices specified by some "educated guess". The resulting quasi-likelihood equations of the demixing matrices take a special form, recently termed an extended "Sequentially Drilled" Joint Congruence (SeDJoCo) transformation, which is reminiscent of (though essentially different from) classical joint diagonalization. We show that asymptotically this QMLE, i.e., the solution of the resulting extended SeDJoCo transformation, attains perfect separation (under some mild conditions) regardless of the sources' true distributions and/or covariance matrices. In addition, based on the "small-errors" assumption, we present a first-order perturbation analysis of the extended SeDJoCo solution. Using the resulting closed-form expressions for the errors in the solution matrices, we provide closed-form expressions for the resulting Interference-to-Source Ratios (ISRs) for IVA. Moreover, we prove that asymptotically the ISRs depend only on the sources' covariances, and not on their specific distributions. As an immediate consequence of this result, we provide an asymptotically attainable lower bound on the resulting ISRs. We also present empirical results, corroborating our analytical derivations, of three simulation experiments concerning two possible model errors - inaccurate covariance matrices and sources' distribution mismodeling.

Improvement of independent vector analysis for closely spaced sources

A Survey of Optimization Methods for Independent Vector Analysis in Audio Source Separation

An Effective Iterative Solution for Independent Vector Analysis with Convergence Guarantees

Adaptive Imaging of Sound Source Based on Total Variation Prior and a Subspace Iteration Integrated Variational Bayesian Method

Vibration sources identification with independent component analysis

Glimpsing Iva: A Framework For Overcomplete/Complete/Undercomplete Convolutive Source Separation

Speech Separation Using Independent Vector Analysis with an Amplitude Variable Gaussian Mixture Model

Switching Independent Vector Analysis and Its Extension to Blind and Spatially Guided Convolutional Beamforming Algorithms

De-noising Vibration Signal of Rotating Machinery with Blind Sources Separation

Rapid Second-Older Blind Source Separation Algorithm for Vibration Signals

Independent Vector Extraction Constrained on Manifold of Half-Length Filters

An Explicit Connection Between Independent Vector Analysis and Tensor Decomposition in Blind Source Separation

Adaptive Beamforming Based on Interference-Plus-Noise Covariance Matrix Reconstruction for Speech Separation

Glimpsing independent vector analysis: Separating more sources than sensors using active and inactive states

The Extended "Sequentially Drilled" Joint Congruence Transformation and its Application in Gaussian Independent Vector Analysis

A novel semi-blind source separation framework towards maximum signal-to-interference ratio

Performance Analysis of the Gaussian Quasi-Maximum Likelihood Approach for Independent Vector Analysis

Independent low-rank matrix analysis based on the Sinkhorn divergence source model for blind source separation

A smart universal single-channel blind source separation method and applications

An Efficient Implementation of Iterative Adaptive Approach for Source Localization.

Independent Vector Analysis with Deep Neural Network Source Priors