Nonlinear blind source separation using kernels.

D Martinez, A Bray
Author Information
  1. D Martinez: LORIA-CNRS, Vandoeuvre-les-Nancy, France.

Abstract

We derive a new method for solving nonlinear blind source separation (BSS) problems by exploiting second-order statistics in a kernel induced feature space. This paper extends a new and efficient closed-form linear algorithm to the nonlinear domain using the kernel trick originally applied in support vector machines (SVMs). This technique could likewise be applied to other linear covariance-based source separation algorithms. Experiments on realistic nonlinear mixtures of speech signals, gas multisensor data, and visual disparity data illustrate the applicability of our approach.

Word Cloud

Created with Highcharts 10.0.0nonlinearsourceseparationnewblindkernellinearusingapplieddataderivemethodsolvingBSSproblemsexploitingsecond-orderstatisticsinducedfeaturespacepaperextendsefficientclosed-formalgorithmdomaintrickoriginallysupportvectormachinesSVMstechniquelikewisecovariance-basedalgorithmsExperimentsrealisticmixturesspeechsignalsgasmultisensorvisualdisparityillustrateapplicabilityapproachNonlinearkernels

Similar Articles

Cited By