El-Hadi Djermoune

El-Hadi Djermoune
University of Lorraine | UdL · CRAN - Centre de Recherche en Automatique de Nancy

PhD

About

75
Publications
5,310
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
367
Citations
Introduction
El-Hadi Djermoune received the Dipl. Ing. degree in electronics from the University of Bejaia, Algeria, in 1998, the M.S. degree in signal processing and control from Université de Lorraine, France, in 1999, and the Ph.D. degree in 2003. He is currently an Associate Professor in the Center Centre de Recherche en Automatique de Nancy, UMR 7039 both from CNRS and Université de Lorraine. His research interests include harmonic retrieval, inverse problems, sparse representation, and hyperspectral image processing.

Publications

Publications (75)
Article
Background: LRP-1 is a multifunctional scavenger receptor belonging to the LDLR family. Due to its capacity to control pericellular levels of various growth factors and proteases, LRP-1 plays a crucial role in membrane proteome dynamics, which appears decisive for tumor progression. Methods: LRP-1 involvement in a TNBC model was assessed using a...
Article
This paper proposes an exact recovery analysis of greedy algorithms for non-negative sparse representations. Orthogonal greedy algorithms such as Orthogonal Matching Pursuit (OMP) and Orthogonal Least Squares (OLS) consist of gradually increasing the solution support and updating the nonzero coefficients in the least squares sense. From a theoretic...
Article
Full-text available
This paper addresses the problem of wood wastes recycling automation. We propose variable selection methods based on near infrared spectroscopic data to select a set of wavebands that captures the main spectral peaks of wood materials to improve the sorting performances. The spectra are first jointly modeled as linear combinations of explanatory va...
Article
Full-text available
Melanoma is the most aggressive form of skin cancer and the most rapidly expanding cancer in terms of worldwide incidence. If primary cutaneous melanoma is mostly treated with a curative wide local excision, malignant melanoma has a poor prognostic and needs other therapeutic approaches. Angiogenesis is a normal physiological process essential in g...
Conference Paper
It is well-known that Orthogonal Matching Pursuit (OMP) recovers the exact support of K-sparse signals under the condition µ < 1/(2K − 1) where µ denotes the mutual coherence of the dictionary. In this communication, we show that under the same condition and if the unknown K-sparse signal is non-negative, the weights of the atoms selected by OMP ar...
Article
Full-text available
Cheyne-Stokes respiration (CSR) is a sleep-disordered breathing characterized by recurrent central apneas alternating with hyperventilation exhibiting a crescendo-decrescendo pattern of tidal volume. This respiration is reported in patients with heart failure, stroke or damage in respiratory centers. It increases mortality for patients with severe...
Article
Full-text available
Orthogonal greedy algorithms are popular sparse signal reconstruction algorithms. Their principle is to select atoms one by one. A series of unconstrained least-square subproblems of gradually increasing size is solved to compute the approximation coefficients, which is efficiently performed using a fast recursive update scheme. When dealing with n...
Conference Paper
Full-text available
La respiration périodique fait partie des troubles respiratoires du sommeil et son association avec l'insuffisance cardiaque constitue un facteur de risque qui conduit à une augmentation de la mortalité. Cet article présente une nouvelle stratégie pour détecter différents niveaux de gravité de la respiration périodique en utilisant un signal d'élec...
Conference Paper
Full-text available
Cette communication concerne la conception, l'implémentation et l'analyse d'algorithmes gloutons pour la reconstruction parci-monieuse sous contrainte de positivité. Ces algorithmes, conçus pour minimiser un critère quadratique sous contraintes de parcimonie et de positivité, généralisent les algorithmes Orthogonal Matching Pursuit et Orthogonal Le...
Conference Paper
Full-text available
Cette communication propose un algorithme de démélange-déconvolution conjoints permettant d'améliorer la résolution des cartes d'abondance des imageurs industriels. L'algorithme est basé sur un critère incluant une régularisation de Tikhonov pour un traitement hors-ligne. Afin de répondre à une forte demande industrielle, l'algorithme est étendu po...
Preprint
Full-text available
Cheyne-Stokes respiration (CSR) is a sleep-disordered breathing characterized by recurrent central apneas alternating with hyperventilation exhibiting a crescendo-decrescendo pattern of tidal volume. This respiration is reported in patients with heart failure, stroke or damage in respiratory centers. It increases mortality for patients with severe...
Conference Paper
Full-text available
Sparse approximation arises in many applications and often leads to a constrained or penalized L0 minimization problem, which was proved to be NP-hard. This paper proposes a revision of existing analyses of NP-hardness of the penalized L0 minimization problem and it introduces a new proof adapted from Natarajan's construction (1995). Moreover, we p...
Conference Paper
Full-text available
It is well-known that Orthogonal Matching Pursuit (OMP) recovers the exact support of K-sparse signals under the condition µ < 1/(2K − 1) where µ denotes the mutual coherence of the dictionary. In this communication, we show that under the same condition and if the unknown K-sparse signal is non-negative, the weights of the atoms selected by OMP ar...
Preprint
Full-text available
Orthogonal greedy algorithms are popular sparse signal reconstruction algorithms. Their principle is to select atoms one by one. A series of unconstrained least-squares subproblems of gradually increasing size is solved to compute the approximation coefficients, which is efficiently performed using a fast recursive update scheme. When dealing with...
Preprint
Full-text available
We bring a contribution to the exact recovery theory of a K-sparse vector in the noiseless setting under the standard condition µ < 1/(2K − 1), where µ denotes the mutual coherence. While it is known that Orthogonal Matching Pursuit (OMP) and Orthogonal Least Squares (OLS) identify the true support in K iterations, we prove that the weights of the...
Article
Full-text available
This paper proposes a hyperspectral image deconvolution algorithm for the online restoration of hyperspectral images as provided by wiskbroom and pushbroom scanning systems. We introduce a least-mean-squares (LMS)-based framework accounting for the convolution kernel noncausality and including nonquadratic (zero attracting and piecewise constant) r...
Conference Paper
Full-text available
The objective of this study is to assess the practical relevance of innovative matrix factorization methods for the preprocessing of long-term ECG. Those signals are generally noisy with complex baseline wander and require preprocessing, such as filtering, to perform a correct analysis. Our goal is to present two innovative algorithms of matrix fac...
Patent
Selon le procédé, - on fait respirer à un sujet (2) de l’air enrichi en CO2 ; - on détecte et on mesure au moins un paramètre représentatif de son activité respiratoire pendant qu'il respire cet air enrichi en CO2 ; et - on enregistre cette mesure au cours du temps et on la stocke comme mesure de référence normale personnalisée de l'activité respir...
Thesis
Full-text available
This manuscript is a synthesis of my research activity at CRAN lab between 2004 and 2017 where my projects dealt with inverse problems in signal and image processing, sparse approximation for harmonic retrieval and simultaneous variable selection, and biological image analysis. The first part of the manuscript is a review of my teaching and researc...
Conference Paper
Full-text available
This paper introduces a framework based on the LMS algorithm for sequential deconvolution of hyperspectral images acquired by industrial pushbroom imaging systems. Considering a sequential model of image blurring phenomenon, we derive a sliding-block zero-attracting LMS algorithm withspectral regularization. The role of each hyperparameter is discu...
Conference Paper
Full-text available
Cet article s'intéresse à la conception d'une méthode séquentielle de déconvolution d'images hyperspectrales acquises par un imageur pushbroom. A partir de l'écriture sous forme séquentielle de l'image floutée, on propose un algorithme de type LMS (least mean squares) par bloc glissant qui inclut des termes de régularisation spatiale et spectrale....
Conference Paper
Full-text available
This article addresses least-squares minimization under sparsity and non-negativity constraints. We propose a recursive implementation of Non-Negative Orthogonal Matching Pursuit (NNOMP) based on the active set method for solving least-squares subproblems. We further propose an improvement of NNOMP, named support-Shrinkage NNOMP (SNNOMP), based on...
Conference Paper
Full-text available
System identification is a data-driven input-output modeling approach more and more used in biology and biomedicine. In this application context, several assays are repeated to estimate the response variability and reproducibility. The inference of the modeling conclusions to the whole population requires to account for the inter-individual variabi...
Article
Full-text available
Estrogen receptor alpha 36 (ERα36) is a variant of the canonical estrogen receptor alpha (ERα66), widely expressed in hormone sensitive cancer cells and whose high expression level correlates with a poor survival prognosis for breast cancer patients. While ERα36 activity have been related to breast cancer progression or acquired resistance to treat...
Article
Full-text available
In this paper, a new method for the estimation of the parameters of multidimensional (R-D) harmonic and damped complex signals in noise is presented. The problem is formulated as R simultaneous sparse approximations of multiple 1-D signals. To get a method able to handle large size signals while maintaining a sufficient resolution, a multigrid dict...
Article
Full-text available
Antiangiogenics are widely used in cancer treatment in combination with chemotherapy and radiotherapy for their vascular effects. Antiangiogenics are supposed to induce morphological and functional changes in the chaotic tumor vasculature that would help enhance the therapeutic efficacy of chemotherapy and radiotherapy through the amelioration of t...
Article
This work aims at studying a method to automatically estimate regularization parameters of non-negative hyperspectral image deconvolution methods. The deconvolution problem is formulated as a multi-objective optimization problem and the properties of the corresponding response surface are studied. Based on these properties, the minimum distance cri...
Conference Paper
This paper presents a new technique for hyperspectral images classification based on simultaneous sparse approximation. The proposed approach consists in formulating the problem as a convex multi-objective optimization problem which incorporates a term favoring the simultaneous sparsity of the estimated coefficients and a term enforcing a regularit...
Conference Paper
Full-text available
System identification is a data-driven input-output modeling approach more and more used in biology and biomedicine. We will take for example pharmacokinetic and pharmacodynamic (PK/PD) studies. In this application context, each assay is always repeated to estimate the response variability and reproductibility. The inference of the modeling conclus...
Conference Paper
Full-text available
This work aims to study a method to automatically estimate regularization parameters of hyperspectral imagesdeconvolution methods. The deconvolution problem is formulated as a multi-objective optimization problem and theproperties of the corresponding response surface are studied. Based on these properties, the minimum distance criterion(MDC) is pr...
Conference Paper
Full-text available
This paper presents a new technique of simultaneous sparse approximation incorporating a regularity constraint along the coefficients matrix rows. This approach is decomposed in two steps: first a sparse representation of the coefficients matrix is obtained using a simultaneous greedy method. Then, a l1 penalty regularization on the derivative of n...
Conference Paper
Full-text available
L'identification de systèmes apparaît de plus en plus dans la modélisation de systèmes biologiques. Dans ce contexte d'application, chaque essai est toujours répété pour estimer la variabilité des réponses. Inférer les résultats à la population nécessite de prendre en compte la variabilité inter-individu dans la procédure de modélisation. Une solut...
Conference Paper
Full-text available
Ce papier présente une nouvelle technique d'approximation parcimonieuse simultanée intégrant une contrainte de régularité le long des lignes de la matrice de coefficients. Cette approche se décompose en deuxétapesdeuxétapes: en premier lieu une représentation parcimonieuse de la matrice des coefficients est obtenue en utilisant une méthode d'approx...
Conference Paper
Full-text available
System identification is a data-driven modeling approach more and more used in biology and biomedicine. In this application context, each assay is always repeated to estimate the response variability. The inference of the modeling conclusions to the whole population requires to account for the inter-individual variability within the modeling proced...
Conference Paper
Full-text available
This paper extends greedy methods to simultaneous sparse approximation. This problem consists in finding good esti-mation of several input signals at once, using different linear combinations of a few elementary signals, drawn from a fixed collection. The sparse algorithms for which simultaneous ver-sions are proposed are namely CoSaMP, OLS and SBR...
Conference Paper
On considère le problème de décomposition d’une séquence de spectres telle qu’on peut en rencontrer en spectroscopie de photo-électrons [16] : les données sont constituées de plusieurs spectres, acquis à des temps différents. Les raies des spectres évoluent lentement et sont modélisées par des gaussiennes d’amplitudes positives. L’objectif de ce tr...
Conference Paper
Full-text available
Geological samples contain variation in compositions even at the scale of 100 µm justifying the use of analytical methods localised at the micrometer scale. This requirement is satisfied by Micro-LIBS tool using a pulsed laser (lambda=266nm, t=5ns) coupled to an optical microscope, allowing a spot analysis around 10µm diameter. However, a single la...
Conference Paper
Full-text available
We propose a sparse modal estimation approach for analyzing 2-D NMR signals. It consists in decomposing the 2-D problem into two 1-D modal estimations. Each 1-D problem is formulated in terms of simultaneous sparse approximation which is efficiently solved using the Simultaneous Orthogonal Matching Pursuit method associated with a multi-grid dictio...
Article
Full-text available
This paper presents a fast time-domain data analysis method for one- and two-dimensional nuclear magnetic resonance (NMR) spectroscopy, assuming Lorentzian lineshapes, based on an adaptive spectral decomposition. The latter is achieved through successive filtering and decimation steps ending up in a decomposition tree. At each node of the tree, the...
Conference Paper
Full-text available
Modal retrieval problem can be addressed using sparse estimation techniques coupled with a multigrid approach. Selection of the initial grid, in the multigrid algorithm, is a critical problem that needs satisfactory solutions. In this paper we propose a strategy for selecting a coarse initial grid which guarantees convergence of the algorithm even...
Chapter
This chapter presents a study of 32 series of acoustic recordings sampled at fs = 2,048 Hz taken using accelerometers placed at the end of the wave guides in the third secondary circuit’s evaporator of the Prototype Fast Reactor (PFR). The first 15 series are injections of water while the remaining recordings are argon injections. Each series is co...
Article
Full-text available
We address the problem of multidimensional modal estimation using sparse estimation techniques coupled with an efficient multigrid approach. Modal dictionaries are obtained by discretizing modal functions (damped complex exponentials). To get a good resolution, it is necessary to choose a fine discretization grid resulting in intractable computatio...
Conference Paper
Full-text available
Methods for subset selection can be used to address the modal retrieval problem using an overcomplete dictionary composed of elementary damped sinusoids. Apart from the related optimization problems, the major difficulty with such techniques is the size of dictionary allowing one to get a sufficient reconstruction error. In this paper, we propose a...
Technical Report
Full-text available
One of the major challenges is using subspace-based approaches for the determination of the parameters of one- or multidimensional signals is their practical applicability on real data containing \emph{damped} exponentials. In this paper, we present a comparison survey of low complexity subspace-based methods and we focus on 2-D nuclear magnetic re...
Article
This paper investigates some issues in physical modeling of metal inert gas/metal active gas (MIG/MAG) welding process in the short arc mode. In this mode, a metal supply is molten in the arc state and then transferred to the weld pool during the short-circuit state. A hybrid model having two distinct continuous states whose switchings are controll...
Article
Full-text available
We present a study of mode variance statistics for three SVD-based estimation methods in the case of a single-mode damped exponential. The methods considered are namely Kumaresan-Tufts, matrix pencil and Kung's direct data approximation. Through first-order perturbation analysis, we derive closed-form expressions of the variance of the complex mode...
Conference Paper
Full-text available
This article presents a statistical analysis of the Matrix Pencil method for estimating the mode and the amplitude of a single damped complex exponential. This study is based on a perturbation analysis of the mode and the amplitude, assuming a high signal-to-noise ratio. Closed-form expressions of the mean and variance of these perturbations are de...
Conference Paper
Full-text available
We propose an adaptive subband decomposition scheme designed to estimate the parameters of two-dimensional (2D) exponential signals from large data sets. The principle of the method consists to perform recursive 2D decimation and estimation steps. At each resulting subband, a stopping rule is evaluated to decide whether the decomposition should be...
Conference Paper
Full-text available
This paper presents a methodology to estimate the parameters of two-dimensional damped/undamped exponentials from high complexity noisy signals, which is the case in 2-D nuclear magnetic resonance spectroscopy signals. The proposed approach performs adaptive subband decomposition combined with a classical frequency estimator based on the Prony mode...
Conference Paper
Full-text available
This paper presents a methodology allowing to estimate the parameters of two-dimensional damped/undamped sinusoids from high complexity noisy signals, which is the case in 2-D nuclear magnetic resonance spectroscopy. The proposed approach performs an adaptive subband decomposition combined with a classical frequency estimator based on the Prony mod...
Conference Paper
Full-text available
In this paper, we present a discrete contour model used to follow the dynamical behaviour of a molten metal bridge in short arc welding. Good results are achieved on experimental movies as bridge edges are well detected. This approach enables the estimation of the relevant variables in the establishment of a model of MIG/MAG welding in short arc mo...
Chapter
Full-text available
For several years, the possibility of using non-iterative high-resolution (HR) spectral estimators instead of Fourier transform (FT) has received considerable attention in the NMR literature. Different approaches have been proposed, including maximum entropy methods, linear prediction (LP) methods, and state space methods. When used in good conditi...
Conference Paper
Full-text available
In this paper, we present a discrete contour model used to fol-low the dynamical behaviour of a molten metal drop in short arc welding. Good results are achieved on experimental movies as drop edges are well detected. This approach enables the esti-mation of the relevant variables in the establishment of a model of MIG/MAG welding in short arc mode...
Conference Paper
Full-text available
In this paper, we present a discrete contour model used to follow the dynamical behaviour of a molten metal drop in short arc welding. Good results are achieved on ex-perimental movies as drop edges are well detected. This approach enables the estimation of the relevant variables in the establishment of a model of MIG/MAG welding in short arc mode.
Conference Paper
Full-text available
Several methods have been developed for estimating the parameters of damped and undamped exponentials in noise, but the performances of such techniques are generally known only in the undamped case. In this paper, we consider two estimation methods: the Kumaresan-Tufts method and the Matrix Pencil approach, and we obtain their estimation performanc...
Conference Paper
Full-text available
In this paper, a parametric spectral estimation method using an adapted filter bank is presented. The subband decomposition is performed classically through filtering and decimation stages. The decision about stopping or carrying on the decomposition of a given node is taken according to a new stopping rule. The latter uses a measure of whiteness o...
Article
This paper presents a non-iterative, fast, and almost automated time-data analysis method for NMR spectroscopy, based on a new adaptive implementation of high resolution methods used in spectral subbands. It is intended to avoid the choice of the decimation factor (or the width of the spectral windows) which, in the case of a uniform decomposition,...
Thesis
Full-text available
Ce travail porte sur l'estimation fréquentielle de sinusoïdes amorties bruitées par décomposition en sous-bandes. Cette approche est préférable pour des signaux de forte complexité (nombreuses composantes, problèmes de résolution dynamique et fréquentielle), ce qui est souvent le cas dans l'application traitée, la spectroscopie RMN.La première cont...
Conference Paper
Full-text available
In this paper, a parametric spectral estimation method using an adaptive subband decomposition approach is presented. The subband decomposition is performed on each band until a measure of whiteness of the resulting residuals is reached. Using Monte Carlo simulations, the results achieved with the proposed method are compared to those obtained with...