Time-Domain Blind Separation of Audio Sources on the Basis of a Complete ICA Decomposition of an Observation Space

Abstract

Abstract-Time-domain algorithms for blind separation of audio sources can be classified as being based either on a partial or complete decomposition of an observation space. The decomposition, especially the complete one, is mostly done under a constraint to reduce the computational burden. However, this constraint potentially restricts the performance. The authors propose a novel time-domain algorithm that is based on a complete unconstrained decomposition of the observation space. The observation space may be defined in a general way, which allows application of long separating filters, although its dimension is low. The decomposition is done by an appropriate independent component analysis (ICA) algorithm giving independent components that are grouped into clusters corresponding to the original sources. Components of the clusters are combined by a reconstruction procedure after estimating microphone responses of the original sources. The authors demonstrate by experiments that the method works effectively with short data, compared to other methods.

Answering six questions in extracting children’s mismatch negativity through combining wavelet decomposition and independent component analysis

“…Furthermore, this study only used FastICA for the decomposition. Recently, more effective algorithms outperforming FastICA have been reported, including, efficient FastICA (EFICA), weights-adjusted second-order blind identification (WASOBI), combination of them (COBI), and approximate joint diagonalization (AJD), and so on (Koldovský et al 2006(Koldovský et al , 2009Koldovský and Tichavský 2011;Tichavsky et al 2008;Tichavsky and Yeredor 2009). It is reasonable to predict that if the performance of ICA decomposition is better in single runs under ICASSO, the finally extracted components by ICASSO should be more reliable.…”

Sparse target cancellation filters with application to semi-blind noise extraction

“…The method derived here is based on the time-domain blind audio separation approach where an observation space is defined; it is decomposed into independent components using Independent Component Analysis (ICA) [24,25]. The observation space is spanned by rows of a data matrix…”

“…, D P . Therefore, blind methods such as the one in [24] select all integer delays from 0 through L where L is the length of the CF. Hence, we consider this blind approach where D i = i−1, i = 1, .…”

A Khatri–Rao subspace approach to blind identification of mixtures of quasi-stationary sources

“…This problem is important, both fundamentally and practically. In terms of applications, the major driving force for the investigation of BID-QSS is blind speech or audio source separation in microphone arrays [2][3][4][5]. The idea of BID-QSS is to utilize the statistically time-varying characteristics of QSSs to identify the unknown system mixing matrix.…”