Well-posedness of the permutation problem in sparse filter estimation with lp minimization
نویسندگان
چکیده
Convolutive source separation is often done in two stages: 1) estimation of the mixing filters and 2) estimation of the sources. Traditional approaches suffer from the ambiguities of arbitrary permutations and scaling in each frequency bin of the estimated filters and/or the sources, and they are usually corrected by taking into account some special properties of the filters/sources. This paper focusses on the filter permutation problem in the absence of scaling, investigating the possible use of the temporal sparsity of the filters as a property enabling permutation correction. Theoretical and experimental results highlight the potential as well as the limits of sparsity as an hypothesis to obtain a well-posed permutation problem. Key-words: sparse filter, convolutive blind source separation, permutation ambiguity, l minimization, Hall’s Marriage Theorem, bi-stochastic matrix ha l-0 06 40 19 8, v er si on 1 10 N ov 2 01 1 Caractère bien posé du problème de permutation pour l’estimation des filtres parcimonieux par minimisation l Résumé : La séparation de source des mélanges convolutifs se fait souvent en deux étapes : 1) estimation des filtres de mélange et 2) estimation des sources. Les approches classiques souffrent d’ambigüıtés de permutation et de facteur d’échelle arbitraire pour chaque fréquence des filtres et/ou des sources estimés. Ces ambigüıtés sont habituellement corrigées en prenant en compte des propriétés particulières des filtres/sources. Cet article se concentre sur le problème de permutation des filtres en l’absence de facteur d’échelle, en explorant l’utilisation potentielle de la parcimonie temporelle des filtres pour résoudre le problème de permutation. Les résultats théoriques et expérimentaux soulignent tant le potentiel que les limites de l’hypothèse de parcimonie pour obtenir un problème bien posé. Mots-clés : Filtres parcimonieux, séparation aveugle de sources, mélange convolutif, ambigüıté de permutation, théorème de mariage de Hall, matrice bi-stochastique ha l-0 06 40 19 8, v er si on 1 10 N ov 2 01 1 Well-posedness of the permutation problem in sparse filter estimation with l minimization 3
منابع مشابه
Sparse representations versus the matched filter
We have considered the problem of detection and estimation of compact sources immersed in a background plus instrumental noise. Sparse approximation to signals deals with the problem of finding a representation of a signal as a linear combination of a small number of elements from a set of signals called dictionary. The estimation of the signal leads to a minimization problem for the amplitude ...
متن کاملFast Reconstruction of SAR Images with Phase Error Using Sparse Representation
In the past years, a number of algorithms have been introduced for synthesis aperture radar (SAR) imaging. However, they all suffer from the same problem: The data size to process is considerably large. In recent years, compressive sensing and sparse representation of the signal in SAR has gained a significant research interest. This method offers the advantage of reducing the sampling rate, bu...
متن کاملLp-norm Regularization Algorithms for Optimization Over Permutation Matrices
Optimization problems over permutation matrices appear widely in facility layout, chip design, scheduling, pattern recognition, computer vision, graph matching, etc. Since this problem is NP-hard due to the combinatorial nature of permutation matrices, we relax the variable to be the more tractable doubly stochastic matrices and add an Lp-norm (0 < p < 1) regularization term to the objective fu...
متن کاملA Multi-objective Immune System for a New Bi-objective Permutation Flowshop Problem with Sequence-dependent Setup Times
We present a new mathematical model for a permutation flowshop scheduling problem with sequence-dependent setup times considering minimization of two objectives, namely makespan and weighted mean total earliness/tardiness. Only small-sized problems with up to 20 jobs can be solved by the proposed integer programming approach. Thus, an effective multi-objective immune system (MOIS) is ...
متن کاملSparsity regularization of the diffusion coefficient problem: well-posedness and convergence rates
In this paper, we investigate sparsity regularization for the diffusion coefficient identification problem. Here, the regularization method is incorporated with the energy functional approach. The advantages of our approach are to deal with convex minimization problems. Therefore, the well-posedness of the problem is obtained without requiring regularity property of the parameter. The convexity...
متن کامل