Comparison of independent component and independent subspace analysis algorithms

Recent advances in separation of convolutive mixtures of audio signals have shown that the problem can be successfully solved in time-domain in a multistep procedure including an application of some method of instantaneous independent component analysis (ICA) or independent subspace analysis (ISA), as one of the steps. In this paper we propose a test that allows a comparison of different ICA and ISA algorithms from this perspective. The test consists in evaluating separation of a pseudo-convolutive mixture of given independent signals. The mixture has features of real-world convolutive mixtures and of instantaneous mixtures simultaneously. We apply the proposed test to compare performance of several ICA and ISA algorithms in four different scenarios, taking in mind that suitability of the algorithms depends on properties of the separated signals.