Source Separation Contrasts Using a Reference Signal

Adib, Abdellah; Moreau, Éric; Aboutajdine, Driss

doi:10.1109/lsp.2003.822894

Cited by 33 publications

(20 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At order , the cost function can also be solved algebraically; indeed, is then given by the dominant eigenvector of the referenceweighted covariance matrix . A similar approach can be found in the QHOC as developed in [14] and [15], where we have in the real, instantaneous case subject to (14) with . In practice, most often a single reference signal is used, which is arbitrarily chosen (e.g., as an arbitrary linear combination of the observations).…”

Section: B Reference-based Filteringmentioning

confidence: 97%

“…Because higher order moments are considered, BSSR offers a better performance when the reference signal has relatively few nonzero values [13]. The quadratic higher order criteria (QHOC) [14], [15] can be considered as a generalization of the BSSR approach. QHOC have been proven to be contrasts for source separation, but they have not been derived in a maximum likelihood sense, so that they do not inherit the estimation optimality of maximum likelihood estimators in the sense of Fisher's information.…”

mentioning

confidence: 99%

See 1 more Smart Citation

Source Extraction by Maximizing the Variance in the Conditional Distribution Tails

Phlypo

Zarzoso

Lemahieu

2010

IEEE Trans. Signal Process.

View full text Add to dashboard Cite

Abstract-This paper presents a method for signal extraction based on conditional second-order moments of the output of the extraction filter. The estimator of the filter is derived from an approximate maximum likelihood criterion conditioned on a presence indicator of the source of interest. The conditional moment is shown to be a contrast function under the conditions that 1) all cross-moments of the same order between the source signal of interest and the other source signals are null and 2) that the source of interest has the largest conditional moment among all sources. For the two-source two-observation case, this allows us to derive the theoretical recovery bounds of the contrast when the conditional cross-moment does not vanish. A comparison with empirical results confirms these bounds. Simulations show that the estimator is quite robust to additive Gaussian distributed noise. Also through simulations, we show that the error level induced by a rough approximation of the presence indicator shows a strong similarity with that of additive noise. The robustness, with respect both to noise and to inaccuracies in the prior information about the source presence, guarantees a wide applicability of the proposed method.

show abstract

Section: B Reference-based Filteringmentioning

confidence: 97%

mentioning

confidence: 99%

Source Extraction by Maximizing the Variance in the Conditional Distribution Tails

Phlypo

Zarzoso

Lemahieu

2010

IEEE Trans. Signal Process.

View full text Add to dashboard Cite

show abstract

“…The mixtures of real-world signals are usually convolutive mixtures, so the conventional methods for the standard blind source separation (BSS) problem which assume instantaneous mixtures of stationary sources [3,4] are no longer appropriate. In a cyclostationary context, BSS has motivated much research which we summarize, for sake of space, as follows.…”

Section: Introductionmentioning

confidence: 99%

On Blind MIMO System Identification Based on Second‐Order Cyclic Statistics

Sabri

Badaoui

Guillet

et al. 2008

Journal of Electrical and Computer Engineering

Self Cite

View full text Add to dashboard Cite

This letter introduces a new frequency domain approach for either MIMO System Identification or Source Separation of convolutive mixtures in cyclostationary context. We apply the joint diagonalization algorithm to a set of cyclic spectral density matrices of the measurements to identify the mixing system at each frequency up to permutation and phase ambiguity matrices. An efficient algorithm to overcome the frequency dependent permutations and to recover the phase, even for non-minimum-phase channels, based on cyclostationarity is also presented. The new approach exploits the fact that each input has a different and specific cyclic frequency. A comparison with an existing MIMO method is proposed.

show abstract

“…From a practical viewpoint, it would be better to estimate the fourth-order cumulant matrixF (t) during t l = 1 to t l all .w [1] i (t) andw [2] i (t) are the eigenvectors obtained for Stage 1) and Stage 2), respectively. t s denotes an arbitrary integer satisfying 2 < t s < t lall .…”

Section: End Endmentioning

confidence: 99%

“…Researches applying this idea to solving blind signal processing (BSP) problems, such as the BD, the BE, the BSS, and so on, have been made by Jelonnek et al (e.g., [5]), Adib et al (e.g., [1]), Rhioui et al [14], and Castella, et al [2]. Jelonnek et al have shown in the single-input case that by the Lagrangian method, the maximization of a contrast function leads to a closed-form solution expressed as a generalized eigenvector problem, which is referred to as an eigenvector algorithm (EVA).…”

Section: Introductionmentioning

confidence: 99%

Blind Deconvolution of MIMO-IIR Systems: A Two-Stage EVA

Kawamoto

Inouye

Kohno

Neural Information Processing

View full text Add to dashboard Cite

Abstract. This paper deals with a blind deconvolution (DB) problem for multiple-input multiple-output infinite impulse response (MIMO-IIR) systems. To solve this problem, we propose an eigenvector algorithm (EVA). In the proposed EVA, two kinds of EVAs are merged so as to give a good performance: One is an EVA and the other is a Robust EVA (REVA) which works with as little sensitive to Gaussian noise as possible. Owing to this combination, two drawbacks of the conventional EVAs can be overcome. Simulation results show the validity of the proposed EVA.

show abstract

Source Separation Contrasts Using a Reference Signal

Cited by 33 publications

References 6 publications

Source Extraction by Maximizing the Variance in the Conditional Distribution Tails

Source Extraction by Maximizing the Variance in the Conditional Distribution Tails

On Blind MIMO System Identification Based on Second‐Order Cyclic Statistics

Blind Deconvolution of MIMO-IIR Systems: A Two-Stage EVA

Contact Info

Product

Resources

About