Moreau, Eric
Overview
Works:  31 works in 54 publications in 2 languages and 898 library holdings 

Genres:  Conference papers and proceedings 
Roles:  Author, Thesis advisor, Other, Opponent, Editor, htt, 956 
Publication Timeline
.
Most widely held works by
Eric Moreau
Blind identification and separation of complexvalued signals by
Eric Moreau(
)
14 editions published in 2013 in English and Undetermined and held by 797 WorldCat member libraries worldwide
Blind identification consists of estimating a multidimensional system only through the use of its output, and source separation, the blind estimation of the inverse of the system. Estimation is generally carried out using different statistics of the output. The authors of this book consider the blind identification and source separation problem in the complexdomain, where the available statistical properties are richer and include noncircularity of the sources  underlying components. They define identifiability conditions and present stateoftheart algorithms that are based on algebraic methods as well as iterative algorithms based on maximum likelihood theory
14 editions published in 2013 in English and Undetermined and held by 797 WorldCat member libraries worldwide
Blind identification consists of estimating a multidimensional system only through the use of its output, and source separation, the blind estimation of the inverse of the system. Estimation is generally carried out using different statistics of the output. The authors of this book consider the blind identification and source separation problem in the complexdomain, where the available statistical properties are richer and include noncircularity of the sources  underlying components. They define identifiability conditions and present stateoftheart algorithms that are based on algebraic methods as well as iterative algorithms based on maximum likelihood theory
Latent variable analysis and signal separation : 9th international conference, LVA/ICA 2010, St. Malo, France, September 2730,
2010 : proceedings by
Vincent Vigneron(
)
6 editions published in 2010 in English and held by 64 WorldCat member libraries worldwide
This book constitutes the proceedings of the 9th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2010, held in St. Malo, France, in September 2010. The 25 papers presented were carefully reviewed and selected from over hundred submissions. The papers collected in this volume demonstrate that the research activity in the field continues to gather theoreticians and practitioners, with contributions ranging range from abstract concepts to the most concrete and applicable questions and considerations. Speech and audio, as well as biomedical applications, continue to carry the mass of the considered applications. Unsurprisingly the concepts of sparsity and nonnegativity, as well as tensor decompositions, have become predominant, reflecting the strong activity on these themes in signal and image processing at large
6 editions published in 2010 in English and held by 64 WorldCat member libraries worldwide
This book constitutes the proceedings of the 9th International Conference on Latent Variable Analysis and Signal Separation, LVA/ICA 2010, held in St. Malo, France, in September 2010. The 25 papers presented were carefully reviewed and selected from over hundred submissions. The papers collected in this volume demonstrate that the research activity in the field continues to gather theoreticians and practitioners, with contributions ranging range from abstract concepts to the most concrete and applicable questions and considerations. Speech and audio, as well as biomedical applications, continue to carry the mass of the considered applications. Unsurprisingly the concepts of sparsity and nonnegativity, as well as tensor decompositions, have become predominant, reflecting the strong activity on these themes in signal and image processing at large
Estimation sous contraintes de communication : algorithmes et performances asymptotiques by
Rodrigo Cabral Farias(
)
2 editions published in 2013 in French and English and held by 3 WorldCat member libraries worldwide
With recent advances in sensing and communication technology, sensor networks have emerged as a new field in signal processing. One of the applications of his new field is remote estimation, where the sensors gather information and send it to some distant point where estimation is carried out. For overcoming the new design challenges brought by this approach (constrained energy, bandwidth and complexity), quantization of the measurements can be considered. Based on this context, we study the problem of estimation based on quantized measurements. We focus mainly on the scalar location parameter estimation problem, the parameter is considered to be either constant or varying according to a slow Wiener process model. We present estimation algorithms to solve this problem and, based on performance analysis, we show the importance of quantizer range adaptiveness for obtaining optimal performance. We propose a low complexity adaptive scheme that jointly estimates the parameter and updates the quantizer thresholds, achieving in this way asymptotically optimal performance. With only 4 or 5 bits of resolution, the asymptotically optimal performance for uniform quantization is shown to be very close to the continuous measurement estimation performance. Finally, we propose a high resolution approach to obtain an approximation of the optimal nonuniform quantization thresholds for parameter estimation and also to obtain an analytical approximation of the estimation performance based on quantized measurements
2 editions published in 2013 in French and English and held by 3 WorldCat member libraries worldwide
With recent advances in sensing and communication technology, sensor networks have emerged as a new field in signal processing. One of the applications of his new field is remote estimation, where the sensors gather information and send it to some distant point where estimation is carried out. For overcoming the new design challenges brought by this approach (constrained energy, bandwidth and complexity), quantization of the measurements can be considered. Based on this context, we study the problem of estimation based on quantized measurements. We focus mainly on the scalar location parameter estimation problem, the parameter is considered to be either constant or varying according to a slow Wiener process model. We present estimation algorithms to solve this problem and, based on performance analysis, we show the importance of quantizer range adaptiveness for obtaining optimal performance. We propose a low complexity adaptive scheme that jointly estimates the parameter and updates the quantizer thresholds, achieving in this way asymptotically optimal performance. With only 4 or 5 bits of resolution, the asymptotically optimal performance for uniform quantization is shown to be very close to the continuous measurement estimation performance. Finally, we propose a high resolution approach to obtain an approximation of the optimal nonuniform quantization thresholds for parameter estimation and also to obtain an analytical approximation of the estimation performance based on quantized measurements
Séparation aveugle de mélanges linéaires convolutifs de sources corrélées by
Hicham Ghennioui(
Book
)
2 editions published in 2008 in French and held by 2 WorldCat member libraries worldwide
Dans cette thèse, nous étudions le problème de la séparation aveugle de mélanges linéaires convolutifs surdéterminés réels ou complexes de sources. Les sources considérées sont réelles ou complexes, déterministes ou aléatoires et dans ce dernier cas statistiquement indépendantes ou corrélées, stationnaires, cyclostationnaires ou nonstationnaires. Nous développons des approches combinant de nouveaux algorithmes de (bloc) diagonalisation conjointe (non unitaires) à de nouveaux détecteurs de points (tempsfréquence ou autres...) particuliers permettant d'élaborer le ou les ensembles de matrices devant être (bloc) diagonalisées conjointement. Les principaux avantages de ces approches sont d'être plus directes en ce se sens qu'elles ne requièrent plus de blanchiment préalable des observations. Elles permettent en outre d'aborder le cas réputé difficile des signaux corrélés. En ce qui concerne les algorithmes de (bloc) diagonalisation conjointe, nous proposons quatre nouveaux algorithmes sans contrainte d'unitarité sur la matrice recherchée. Le premier algorithme est de type algébrique itératif. Il est basé sur l'optimisation d'un critère de type moindres carrés. Les trois autres approches utilisent un schéma d'optimisation de type gradient. Dans un premier temps le calcul du gradient matriciel de la fonction de coût étudiée est approché. Puis dans un second temps le calcul exact est mené et deux nouveaux algorithmes sont proposés : l'un à base de gradient, l'autre à base de gradient relatif. Nous étudions les versions à pas fixe de ces trois algorithmes, puis les versions à pas optimal afin d'accélérer la convergence des algorithmes (le pas est alors recalculé algébriquement à chaque itération en cherchant les racines d'un polynôme d'ordre trois). Un lien avec la diagonalisation conjointe non unitaire est également établi. Ces algorithmes de blocdiagonalisation conjointe possèdent l'avantage d'être généraux : les matrices de l'ensemble considéré ne sont ni nécessairement réelles, ni à symétrie hermitienne, ni définies positives et le blocdiagonaliseur conjoint peut être une matrice unitaire ou nonunitaire
2 editions published in 2008 in French and held by 2 WorldCat member libraries worldwide
Dans cette thèse, nous étudions le problème de la séparation aveugle de mélanges linéaires convolutifs surdéterminés réels ou complexes de sources. Les sources considérées sont réelles ou complexes, déterministes ou aléatoires et dans ce dernier cas statistiquement indépendantes ou corrélées, stationnaires, cyclostationnaires ou nonstationnaires. Nous développons des approches combinant de nouveaux algorithmes de (bloc) diagonalisation conjointe (non unitaires) à de nouveaux détecteurs de points (tempsfréquence ou autres...) particuliers permettant d'élaborer le ou les ensembles de matrices devant être (bloc) diagonalisées conjointement. Les principaux avantages de ces approches sont d'être plus directes en ce se sens qu'elles ne requièrent plus de blanchiment préalable des observations. Elles permettent en outre d'aborder le cas réputé difficile des signaux corrélés. En ce qui concerne les algorithmes de (bloc) diagonalisation conjointe, nous proposons quatre nouveaux algorithmes sans contrainte d'unitarité sur la matrice recherchée. Le premier algorithme est de type algébrique itératif. Il est basé sur l'optimisation d'un critère de type moindres carrés. Les trois autres approches utilisent un schéma d'optimisation de type gradient. Dans un premier temps le calcul du gradient matriciel de la fonction de coût étudiée est approché. Puis dans un second temps le calcul exact est mené et deux nouveaux algorithmes sont proposés : l'un à base de gradient, l'autre à base de gradient relatif. Nous étudions les versions à pas fixe de ces trois algorithmes, puis les versions à pas optimal afin d'accélérer la convergence des algorithmes (le pas est alors recalculé algébriquement à chaque itération en cherchant les racines d'un polynôme d'ordre trois). Un lien avec la diagonalisation conjointe non unitaire est également établi. Ces algorithmes de blocdiagonalisation conjointe possèdent l'avantage d'être généraux : les matrices de l'ensemble considéré ne sont ni nécessairement réelles, ni à symétrie hermitienne, ni définies positives et le blocdiagonaliseur conjoint peut être une matrice unitaire ou nonunitaire
Traitements spatiotemporels adaptés aux radars bistatiques à émetteurs non coopératifs by
Jacques Raout(
Book
)
2 editions published in 2010 in French and held by 2 WorldCat member libraries worldwide
The framework of this research is bistatic radars using noncooperative transmitters. The study focuses on digital video broadcasting transmitters. Considering such transmitters leads to higher capabilities of detection compared to other kind of radio communication illuminators. The processing of real data collected from a fixed receiver in the case of different kind of targets (helicopter and boat) leads to a specific method of clutter rejection and target detection. It also provides precious information on statistical properties of the environment and allows to adapt spacetime processing to noiselike signals. The following points are presented in the case of a fixe receiver :  clutter rejection and target detection method in the case of noiselike signals  validation on real DVBT signals  determination of statistical properties of clutter; in the case of a mobile receiver :  adaptation of existing spacetime processing methods and comparison of their performances  adaptation and generalization to the case of multiple targets of the hybridization of both a method of reduced dimension and a non statistical one focusing on the range cell under test  development of a new method of iterative clutter rejection and target detection
2 editions published in 2010 in French and held by 2 WorldCat member libraries worldwide
The framework of this research is bistatic radars using noncooperative transmitters. The study focuses on digital video broadcasting transmitters. Considering such transmitters leads to higher capabilities of detection compared to other kind of radio communication illuminators. The processing of real data collected from a fixed receiver in the case of different kind of targets (helicopter and boat) leads to a specific method of clutter rejection and target detection. It also provides precious information on statistical properties of the environment and allows to adapt spacetime processing to noiselike signals. The following points are presented in the case of a fixe receiver :  clutter rejection and target detection method in the case of noiselike signals  validation on real DVBT signals  determination of statistical properties of clutter; in the case of a mobile receiver :  adaptation of existing spacetime processing methods and comparison of their performances  adaptation and generalization to the case of multiple targets of the hybridization of both a method of reduced dimension and a non statistical one focusing on the range cell under test  development of a new method of iterative clutter rejection and target detection
Nouvelles approches pour la séparation de sources by
Saloua Rhioui(
Book
)
2 editions published in 2006 in French and held by 2 WorldCat member libraries worldwide
The blind source separation is widely studied problem in the community of signal processing because of its numerous potential applications in telecommunications, biomedical, geophysics, speech processing, image processing, radar, sonar, etc. This partly explains the great popularity of this set of themes. This thesis proposes new approaches for random sources separation. Two major problems are tackled. The first relates to the instantaneous mixture separation of cyclostationnary sources and the second, more consequent, adresses the problem of the separation of convolutive mixtures stationary signals. He develop new criteria and algorithms based on the use of various orders statistics. More precisely we propose a separation method based on the decomposition of a linear operator combined with a classification method. The performances are compared by computer simulations with PARAFAC decomposition method.Then we are interested in new contrast functions using reference signals within the framework of convolutive mixtures. A MISO contrast is proposed whose great advantage is to allow a quadratic optimization. Then MIMO contrasts are proposed generalizing the existing contrasts in the literature
2 editions published in 2006 in French and held by 2 WorldCat member libraries worldwide
The blind source separation is widely studied problem in the community of signal processing because of its numerous potential applications in telecommunications, biomedical, geophysics, speech processing, image processing, radar, sonar, etc. This partly explains the great popularity of this set of themes. This thesis proposes new approaches for random sources separation. Two major problems are tackled. The first relates to the instantaneous mixture separation of cyclostationnary sources and the second, more consequent, adresses the problem of the separation of convolutive mixtures stationary signals. He develop new criteria and algorithms based on the use of various orders statistics. More precisely we propose a separation method based on the decomposition of a linear operator combined with a classification method. The performances are compared by computer simulations with PARAFAC decomposition method.Then we are interested in new contrast functions using reference signals within the framework of convolutive mixtures. A MISO contrast is proposed whose great advantage is to allow a quadratic optimization. Then MIMO contrasts are proposed generalizing the existing contrasts in the literature
Méthodes pour l'électroencéphalographie multisujet et application aux interfaces cerveauordinateur by
Louis Korczowski(
)
1 edition published in 2018 in French and held by 2 WorldCat member libraries worldwide
The study of several brains interacting (hyperscanning) with neuroimagery allows to extend our understanding of social neurosciences. We propose a framework for hyperscanning using multiuser BrainComputer Interfaces (BCI) that includes several social paradigms such as cooperation or competition. This dissertation includes three interdependent contribution. The first contribution is the development of an experimental platform consisting of a multiplayer video game, namely Brain Invaders 2, controlled by classification of visual event related potentials (ERP) recorded by electroencephalography (EEG). The plateform is validated through two experimental protocols including nineteen and twenty two pairs of subjects while using different adaptive classification approaches using Riemannian geometry. Those approaches are theoretically and experimentally compared during the second contribution ; we demonstrates the superiority in term of accuracy of merging independent classifications over the classification of the hyperbrain during the second contribution. Analysis of interbrain synchronizations is a common approach for hyperscanning, however it is challenging for transient EEG waves with an great spatiotemporal variability (intra and intersubject) and with low signaltonoise ratio such as ERP. Therefore, as third contribution, we propose a new blind source separation model, namely composite model, to extract simultaneously evoked EEG sources and ongoing EEG sources that allows to compensate this variability. A solution using approximate joint diagonalization is given and implemented with a fast Jacobilike algorithm. We demonstrate on Brain Invaders 2 data that our solution extracts simultaneously evoked and ongoing EEG sources and performs better in term of accuracy and robustness compared to the existing models
1 edition published in 2018 in French and held by 2 WorldCat member libraries worldwide
The study of several brains interacting (hyperscanning) with neuroimagery allows to extend our understanding of social neurosciences. We propose a framework for hyperscanning using multiuser BrainComputer Interfaces (BCI) that includes several social paradigms such as cooperation or competition. This dissertation includes three interdependent contribution. The first contribution is the development of an experimental platform consisting of a multiplayer video game, namely Brain Invaders 2, controlled by classification of visual event related potentials (ERP) recorded by electroencephalography (EEG). The plateform is validated through two experimental protocols including nineteen and twenty two pairs of subjects while using different adaptive classification approaches using Riemannian geometry. Those approaches are theoretically and experimentally compared during the second contribution ; we demonstrates the superiority in term of accuracy of merging independent classifications over the classification of the hyperbrain during the second contribution. Analysis of interbrain synchronizations is a common approach for hyperscanning, however it is challenging for transient EEG waves with an great spatiotemporal variability (intra and intersubject) and with low signaltonoise ratio such as ERP. Therefore, as third contribution, we propose a new blind source separation model, namely composite model, to extract simultaneously evoked EEG sources and ongoing EEG sources that allows to compensate this variability. A solution using approximate joint diagonalization is given and implemented with a fast Jacobilike algorithm. We demonstrate on Brain Invaders 2 data that our solution extracts simultaneously evoked and ongoing EEG sources and performs better in term of accuracy and robustness compared to the existing models
Techniques tensorielles pour le traitement du signal : algorithmes pour la décomposition polyadique canonique by
Alex Pereira da Silva(
)
1 edition published in 2016 in English and held by 2 WorldCat member libraries worldwide
Low rank tensor decomposition has been playing for the last years an important rolein many applications such as blind source separation, telecommunications, sensor array processing,neuroscience, chemometrics, and data mining. The Canonical Polyadic tensor decomposition is veryattractive when compared to standard matrixbased tools, manly on system identification. In this thesis,we propose: (i) several algorithms to compute specific low rankapproximations: finite/iterativerank1 approximations, iterative deflation approximations, and orthogonal tensor decompositions. (ii)A new strategy to solve multivariate quadratic systems, where this problem is reduced to a best rank1 tensor approximation problem. (iii) Theoretical results to study and proof the performance or theconvergence of some algorithms. All performances are supported by numerical experiments
1 edition published in 2016 in English and held by 2 WorldCat member libraries worldwide
Low rank tensor decomposition has been playing for the last years an important rolein many applications such as blind source separation, telecommunications, sensor array processing,neuroscience, chemometrics, and data mining. The Canonical Polyadic tensor decomposition is veryattractive when compared to standard matrixbased tools, manly on system identification. In this thesis,we propose: (i) several algorithms to compute specific low rankapproximations: finite/iterativerank1 approximations, iterative deflation approximations, and orthogonal tensor decompositions. (ii)A new strategy to solve multivariate quadratic systems, where this problem is reduced to a best rank1 tensor approximation problem. (iii) Theoretical results to study and proof the performance or theconvergence of some algorithms. All performances are supported by numerical experiments
Modélisation et analyse de la parole : Contrôle d'un robot parlant via un modèle interne optimal basé sur les réseaux
de neurones artificiels. Outils statistiques en analyse de la parole. by
Iaroslav Blagouchine(
)
2 editions published in 2010 in French and held by 2 WorldCat member libraries worldwide
This Ph.D. dissertation deals with speech modeling and processing, which both share the speech quality aspect. An optimum internal model with constraints is proposed and discussed for the control of a biomechanical speech robot based on the equilibrium point hypothesis (EPH, lambdamodel). It is supposed that the robot internal space is composed of the motor commands lambda of the equilibrium point hypothesis. The main idea of the work is that the robot movements, and in particular the robot speech production, are carried out in such a way that, the length of the path, traveled in the internal space, is minimized under acoustical and mechanical constraints. Mathematical aspect of the problem leads to one of the problems of variational calculus, the socalled geodesic problem, whose exact analytical solution is quite complicated. By using some empirical findings, an approximate solution for the proposed optimum internal model is then developed and implemented. It gives interesting and challenging results, and shows that the proposed internal model is quite realistic; namely, some similarities are found between the robot speech and the real one. Next, by aiming to analyze speech signals, several methods of statistical speech signal processing are developed. They are based on higherorder statistics (namely, on normalized central moments and on the fourthorder cumulant), as well as on the discrete normalized entropy. In this framework, we also designed an unbiased and efficient estimator of the fourthorder cumulant in both batch and adaptive versions
2 editions published in 2010 in French and held by 2 WorldCat member libraries worldwide
This Ph.D. dissertation deals with speech modeling and processing, which both share the speech quality aspect. An optimum internal model with constraints is proposed and discussed for the control of a biomechanical speech robot based on the equilibrium point hypothesis (EPH, lambdamodel). It is supposed that the robot internal space is composed of the motor commands lambda of the equilibrium point hypothesis. The main idea of the work is that the robot movements, and in particular the robot speech production, are carried out in such a way that, the length of the path, traveled in the internal space, is minimized under acoustical and mechanical constraints. Mathematical aspect of the problem leads to one of the problems of variational calculus, the socalled geodesic problem, whose exact analytical solution is quite complicated. By using some empirical findings, an approximate solution for the proposed optimum internal model is then developed and implemented. It gives interesting and challenging results, and shows that the proposed internal model is quite realistic; namely, some similarities are found between the robot speech and the real one. Next, by aiming to analyze speech signals, several methods of statistical speech signal processing are developed. They are based on higherorder statistics (namely, on normalized central moments and on the fourthorder cumulant), as well as on the discrete normalized entropy. In this framework, we also designed an unbiased and efficient estimator of the fourthorder cumulant in both batch and adaptive versions
Algorithmes pour la diagonalisation conjointe de tenseurs sans contrainte unitaire. Application à la séparation MIMO de
sources de télécommunications numériques by
Victor Maurandi(
)
1 edition published in 2015 in French and held by 1 WorldCat member library worldwide
This thesis develops joint diagonalization of matrices and thirdorder tensors methods for MIMO source separation in the field of digital telecommunications. After a state of the art, the motivations and the objectives are presented. Then the joint diagonalisation and the blind source separation issues are defined and a link between both fields is established. Thereafter, five Jacobilike iterative algorithms based on an LU parameterization are developed. For each of them, we propose to derive the diagonalization matrix by optimizing an inverse criterion. Two ways are investigated : minimizing the criterion in a direct way or assuming that the elements from the considered set are almost diagonal. Regarding the parameters derivation, two strategies are implemented : one consists in estimating each parameter independently, the other consists in the independent derivation of couple of wellchosen parameters. Hence, we propose three algorithms for the joint diagonalization of symmetric complex matrices or hermitian ones. The first one relies on searching for the roots of the criterion derivative, the second one relies on a minor eigenvector research and the last one relies on a gradient descent method enhanced by computation of the optimal adaptation step. In the framework of joint diagonalization of symmetric, INDSCAL or non symmetric thirdorder tensors, we have developed two algorithms. For each of them, the parameters derivation is done by computing the roots of the considered criterion derivative. We also show the link between the joint diagonalization of a thirdorder tensor set and the canonical polyadic decomposition of a fourthorder tensor. We confront both methods through numerical simulations. The good behavior of the proposed algorithms is illustrated by means of computing simulations. Finally, they are applied to the source separation of digital telecommunication signals
1 edition published in 2015 in French and held by 1 WorldCat member library worldwide
This thesis develops joint diagonalization of matrices and thirdorder tensors methods for MIMO source separation in the field of digital telecommunications. After a state of the art, the motivations and the objectives are presented. Then the joint diagonalisation and the blind source separation issues are defined and a link between both fields is established. Thereafter, five Jacobilike iterative algorithms based on an LU parameterization are developed. For each of them, we propose to derive the diagonalization matrix by optimizing an inverse criterion. Two ways are investigated : minimizing the criterion in a direct way or assuming that the elements from the considered set are almost diagonal. Regarding the parameters derivation, two strategies are implemented : one consists in estimating each parameter independently, the other consists in the independent derivation of couple of wellchosen parameters. Hence, we propose three algorithms for the joint diagonalization of symmetric complex matrices or hermitian ones. The first one relies on searching for the roots of the criterion derivative, the second one relies on a minor eigenvector research and the last one relies on a gradient descent method enhanced by computation of the optimal adaptation step. In the framework of joint diagonalization of symmetric, INDSCAL or non symmetric thirdorder tensors, we have developed two algorithms. For each of them, the parameters derivation is done by computing the roots of the considered criterion derivative. We also show the link between the joint diagonalization of a thirdorder tensor set and the canonical polyadic decomposition of a fourthorder tensor. We confront both methods through numerical simulations. The good behavior of the proposed algorithms is illustrated by means of computing simulations. Finally, they are applied to the source separation of digital telecommunication signals
Utilisation des Copules en Séparation Aveugle de Sources Indépendantes/Dépendantes by
Nezha Mamouni(
)
1 edition published in 2020 in French and held by 1 WorldCat member library worldwide
The problem of Blind Source Separation (BSS) consists in retrieving unobserved mixed signals from unknown mixtures of them, where there is no, or very limited, information about the source signals and/or the mixing system. In this thesis, we present algorithms in order to separate instantaneous and convolutive mixtures. The principle of these algorithms is to minimize, appropriate separation criteria based on copula densities, using descent gradient type algorithms. These methods can magnificently separate instantaneous and convolutive mixtures of possibly dependent source components even when the copula model is unknown
1 edition published in 2020 in French and held by 1 WorldCat member library worldwide
The problem of Blind Source Separation (BSS) consists in retrieving unobserved mixed signals from unknown mixtures of them, where there is no, or very limited, information about the source signals and/or the mixing system. In this thesis, we present algorithms in order to separate instantaneous and convolutive mixtures. The principle of these algorithms is to minimize, appropriate separation criteria based on copula densities, using descent gradient type algorithms. These methods can magnificently separate instantaneous and convolutive mixtures of possibly dependent source components even when the copula model is unknown
Traitements spatiotemporels adaptés aux radars bistatiques à émetteurs non coopératifs by
Jacques Raout(
)
1 edition published in 2010 in French and held by 1 WorldCat member library worldwide
The framework of this research is bistatic radars using noncooperative transmitters. The study focuses on digital video broadcasting transmitters. Considering such transmitters leads to higher capabilities of detection compared to other kind of radio communication illuminators. The processing of real data collected from a fixed receiver in the case of different kind of targets (helicopter and boat) leads to a specific method of clutter rejection and target detection. It also provides precious information on statistical properties of the environment and allows to adapt spacetime processing to noiselike signals. The following points are presented in the case of a fixe receiver :  clutter rejection and target detection method in the case of noiselike signals  validation on real DVBT signals  determination of statistical properties of clutter; in the case of a mobile receiver :  adaptation of existing spacetime processing methods and comparison of their performances  adaptation and generalization to the case of multiple targets of the hybridization of both a method of reduced dimension and a non statistical one focusing on the range cell under test  development of a new method of iterative clutter rejection and target detection
1 edition published in 2010 in French and held by 1 WorldCat member library worldwide
The framework of this research is bistatic radars using noncooperative transmitters. The study focuses on digital video broadcasting transmitters. Considering such transmitters leads to higher capabilities of detection compared to other kind of radio communication illuminators. The processing of real data collected from a fixed receiver in the case of different kind of targets (helicopter and boat) leads to a specific method of clutter rejection and target detection. It also provides precious information on statistical properties of the environment and allows to adapt spacetime processing to noiselike signals. The following points are presented in the case of a fixe receiver :  clutter rejection and target detection method in the case of noiselike signals  validation on real DVBT signals  determination of statistical properties of clutter; in the case of a mobile receiver :  adaptation of existing spacetime processing methods and comparison of their performances  adaptation and generalization to the case of multiple targets of the hybridization of both a method of reduced dimension and a non statistical one focusing on the range cell under test  development of a new method of iterative clutter rejection and target detection
Blind identification and separation of complexvalued signals by
Eric Moreau(
)
1 edition published in 2013 in English and held by 1 WorldCat member library worldwide
1 edition published in 2013 in English and held by 1 WorldCat member library worldwide
Breaking the curse of dimensionality based on tensor train : models and algorithms by
Yassine Zniyed(
)
1 edition published in 2019 in English and held by 1 WorldCat member library worldwide
Massive and heterogeneous data processing and analysis have been clearly identified by the scientific community as key problems in several application areas. It was popularized under the generic terms of "data science" or "big data". Processing large volumes of data, extracting their hidden patterns, while preforming prediction and inference tasks has become crucial in economy, industry and science.Treating independently each set of measured data is clearly a reductiveapproach. By doing that, "hidden relationships" or intercorrelations between thedatasets may be totally missed. Tensor decompositions have received a particular attention recently due to their capability to handle a variety of mining tasks applied to massive datasets, being a pertinent framework taking into account the heterogeneity and multimodality of the data. In this case, data can be arranged as a Ddimensional array, also referred to as a Dorder tensor.In this context, the purpose of this work is that the following properties are present: (i) having a stable factorization algorithms (not suffering from convergence problems), (ii) having a low storage cost (i.e., the number of free parameters must be linear in D), and (iii) having a formalism in the form of a graph allowing a simple but rigorous mental visualization of tensor decompositions of tensors of high order, i.e., for D> 3.Therefore, we rely on the tensor train decomposition (TT) to develop new TT factorization algorithms, and new equivalences in terms of tensor modeling, allowing a new strategy of dimensionality reduction and criterion optimization of coupled least squares for the estimation of parameters named JIRAFE.This methodological work has had applications in the context of multidimensional spectral analysis and relay telecommunications systems
1 edition published in 2019 in English and held by 1 WorldCat member library worldwide
Massive and heterogeneous data processing and analysis have been clearly identified by the scientific community as key problems in several application areas. It was popularized under the generic terms of "data science" or "big data". Processing large volumes of data, extracting their hidden patterns, while preforming prediction and inference tasks has become crucial in economy, industry and science.Treating independently each set of measured data is clearly a reductiveapproach. By doing that, "hidden relationships" or intercorrelations between thedatasets may be totally missed. Tensor decompositions have received a particular attention recently due to their capability to handle a variety of mining tasks applied to massive datasets, being a pertinent framework taking into account the heterogeneity and multimodality of the data. In this case, data can be arranged as a Ddimensional array, also referred to as a Dorder tensor.In this context, the purpose of this work is that the following properties are present: (i) having a stable factorization algorithms (not suffering from convergence problems), (ii) having a low storage cost (i.e., the number of free parameters must be linear in D), and (iii) having a formalism in the form of a graph allowing a simple but rigorous mental visualization of tensor decompositions of tensors of high order, i.e., for D> 3.Therefore, we rely on the tensor train decomposition (TT) to develop new TT factorization algorithms, and new equivalences in terms of tensor modeling, allowing a new strategy of dimensionality reduction and criterion optimization of coupled least squares for the estimation of parameters named JIRAFE.This methodological work has had applications in the context of multidimensional spectral analysis and relay telecommunications systems
Séparation de sources et fonctions de contraste by
Eric Moreau(
Book
)
1 edition published in 2000 in French and held by 1 WorldCat member library worldwide
1 edition published in 2000 in French and held by 1 WorldCat member library worldwide
Décompositions Matricielles Conjointes et Séparation Aveugle de Sources by
El Hostafa Fadaili(
)
1 edition published in 2006 in French and held by 1 WorldCat member library worldwide
1 edition published in 2006 in French and held by 1 WorldCat member library worldwide
Latent variable analysis and signal separation(
)
1 edition published in 2010 in English and held by 1 WorldCat member library worldwide
1 edition published in 2010 in English and held by 1 WorldCat member library worldwide
Algorithmes de diagonalisation conjointe par similitude pour la décomposition canonique polyadique de tenseurs : applications
en séparation de sources by
Rémi André(
)
1 edition published in 2018 in French and held by 1 WorldCat member library worldwide
This thesis introduces new joint eigenvalue decomposition algorithms. These algorithms allowamongst others to solve the canonical polyadic decomposition problem. This decomposition iswidely used for blind source separation. Using the joint eigenvalue decomposition to solve thecanonical polyadic decomposition problem allows to avoid some problems whose the others canonicalpolyadic decomposition algorithms generally suffer, such as the convergence rate, theoverfactoring sensibility and the correlated factors sensibility. The joint eigenvalue decompositionalgorithms dealing with complex data give either good results when the noise power is low, orthey are robust to the noise power but have a high numerical cost. Therefore, we first proposealgorithms equally dealing with real and complex. Moreover, in some applications, factor matricesof the canonical polyadic decomposition contain only nonnegative values. Taking this constraintinto account makes the algorithms more robust to the overfactoring and to the correlated factors.Therefore, we also offer joint eigenvalue decomposition algorithms taking advantage of thisnonnegativity constraint. Suggested numerical simulations show that the first developed algorithmsimprove the estimation accuracy and reduce the numerical cost in the case of complexdata. Our numerical simulations also highlight the fact that our nonnegative joint eigenvaluedecomposition algorithms improve the factor matrices estimation when their columns have ahigh correlation degree. Eventually, we successfully applied our algorithms to two blind sourceseparation problems : one concerning numerical telecommunications and the other concerningfluorescence spectroscopy
1 edition published in 2018 in French and held by 1 WorldCat member library worldwide
This thesis introduces new joint eigenvalue decomposition algorithms. These algorithms allowamongst others to solve the canonical polyadic decomposition problem. This decomposition iswidely used for blind source separation. Using the joint eigenvalue decomposition to solve thecanonical polyadic decomposition problem allows to avoid some problems whose the others canonicalpolyadic decomposition algorithms generally suffer, such as the convergence rate, theoverfactoring sensibility and the correlated factors sensibility. The joint eigenvalue decompositionalgorithms dealing with complex data give either good results when the noise power is low, orthey are robust to the noise power but have a high numerical cost. Therefore, we first proposealgorithms equally dealing with real and complex. Moreover, in some applications, factor matricesof the canonical polyadic decomposition contain only nonnegative values. Taking this constraintinto account makes the algorithms more robust to the overfactoring and to the correlated factors.Therefore, we also offer joint eigenvalue decomposition algorithms taking advantage of thisnonnegativity constraint. Suggested numerical simulations show that the first developed algorithmsimprove the estimation accuracy and reduce the numerical cost in the case of complexdata. Our numerical simulations also highlight the fact that our nonnegative joint eigenvaluedecomposition algorithms improve the factor matrices estimation when their columns have ahigh correlation degree. Eventually, we successfully applied our algorithms to two blind sourceseparation problems : one concerning numerical telecommunications and the other concerningfluorescence spectroscopy
Apprentissage profond et traitement d'images pour la détection de fumée by
Rabeb Kaabi(
)
1 edition published in 2020 in French and held by 1 WorldCat member library worldwide
This thesis deals with the problem of forest fire detection using image processing and machine learning tools. A forest fire is a fire that spreads over a wooded area. It can be of natural origin (due to lightning or a volcanic eruption) or human. Around the world, the impact of forest fires on many aspects of our daily lives is becoming more and more apparent on the entire ecosystem.Many methods have been shown to be effective in detecting forest fires. The originality of the present work lies in the early detection of fires through the detection of forest smoke and the classification of smoky and nonsmoky regions using deep learning and image processing tools. A set of preprocessing techniques helped us to have an important database which allowed us afterwards to test the robustness of the model based on deep belief network we proposed and to evaluate the performance by calculating the following metrics (IoU, Accuracy, Recall, F1 score). Finally, the proposed algorithm is tested on several images in order to validate its efficiency. The simulations of our algorithm have been compared with those processed in the state of the art (Deep CNN, SVM...) and have provided very good results. The results of the proposed methods gave an average classification accuracy of about 96.5% for the early detection of smoke
1 edition published in 2020 in French and held by 1 WorldCat member library worldwide
This thesis deals with the problem of forest fire detection using image processing and machine learning tools. A forest fire is a fire that spreads over a wooded area. It can be of natural origin (due to lightning or a volcanic eruption) or human. Around the world, the impact of forest fires on many aspects of our daily lives is becoming more and more apparent on the entire ecosystem.Many methods have been shown to be effective in detecting forest fires. The originality of the present work lies in the early detection of fires through the detection of forest smoke and the classification of smoky and nonsmoky regions using deep learning and image processing tools. A set of preprocessing techniques helped us to have an important database which allowed us afterwards to test the robustness of the model based on deep belief network we proposed and to evaluate the performance by calculating the following metrics (IoU, Accuracy, Recall, F1 score). Finally, the proposed algorithm is tested on several images in order to validate its efficiency. The simulations of our algorithm have been compared with those processed in the state of the art (Deep CNN, SVM...) and have provided very good results. The results of the proposed methods gave an average classification accuracy of about 96.5% for the early detection of smoke
Décompositions conjointes de matrices complexes : application à la séparation de sources by
Tual Trainini(
)
1 edition published in 2012 in French and held by 1 WorldCat member library worldwide
This thesis deals with the study of joint diagonalization of complex matrices methods for source separation, wether in the field of numerical telecommunications and radioastronomy. After having introduced the motivations that drove this study, we present a brief stateoftheart in the field. The joint diagonalization and source separation problems are reminded, and a link between these two themes is established. Thereafter, several iterative algorithms are developed. First, methods using a gradientlike update of the separation matrix are introduced. They are based on wise approximations of the considered criterion. In order to improve the convergence speed, a method using a computation of an optimal step size is presented, and variations around this computation, based on the previously introduced approximations are done. Two other approaches are then introduced. The first one analytically determines the separation matrix, by algebraically computing the terms composing the update matrix pairwise from a linear equation system. The second one recursively estimates the mixing matrix, based on an alternating least squares method. In order to enhance the convergence speed, a seek of an enhanced line search algorithm is proposed. These methods are then validated on a classical joint diagonalization problem. Aterwards, these algorithms are applied to the source separation of numerical communication signals, while using second or higher order statistics. Comparisons are also made with wellknown methods. The second application relates to elimination of rterrestrial interferences from the estimation of the associated space in order to observe at best cosmic sources from LOFAR station data
1 edition published in 2012 in French and held by 1 WorldCat member library worldwide
This thesis deals with the study of joint diagonalization of complex matrices methods for source separation, wether in the field of numerical telecommunications and radioastronomy. After having introduced the motivations that drove this study, we present a brief stateoftheart in the field. The joint diagonalization and source separation problems are reminded, and a link between these two themes is established. Thereafter, several iterative algorithms are developed. First, methods using a gradientlike update of the separation matrix are introduced. They are based on wise approximations of the considered criterion. In order to improve the convergence speed, a method using a computation of an optimal step size is presented, and variations around this computation, based on the previously introduced approximations are done. Two other approaches are then introduced. The first one analytically determines the separation matrix, by algebraically computing the terms composing the update matrix pairwise from a linear equation system. The second one recursively estimates the mixing matrix, based on an alternating least squares method. In order to enhance the convergence speed, a seek of an enhanced line search algorithm is proposed. These methods are then validated on a classical joint diagonalization problem. Aterwards, these algorithms are applied to the source separation of numerical communication signals, while using second or higher order statistics. Comparisons are also made with wellknown methods. The second application relates to elimination of rterrestrial interferences from the estimation of the associated space in order to observe at best cosmic sources from LOFAR station data
more
fewer
Audience Level
0 

1  
Kids  General  Special 
Related Identities
 Adali, Tülay
 Gribonval, Rémi Editor
 Zarzoso, Vicente Opponent Editor
 Vigneron, Vincent Author Editor
 Vincent, Emmanuel Editor
 Université de Toulon Degree grantor
 Université du Sud ToulonVar UFR de Sciences et Techniques Other Degree grantor
 École doctorale électronique, électrotechnique, automatique, traitement du signal (Grenoble) Other
 Grenoble Images parole signal automatique Other
 ThirionMoreau, Nadège (19......). Thesis advisor
Useful Links
Associated Subjects
Computational complexity Computer science Computer simulation Computer software Computer vision Electronic noise MIMO systems Multivariate analysis Neural networks (Computer science) Optical pattern recognition Signal processing Signal processingDigital techniques Signal processingStatistical methods Software engineering Source separation (Signal processing)