O objetivo da desconvolução cega de imagens é restaurar uma imagem degradada sem usar informação da imagem real ou da função de degradação. O mapeamento dos níveis de cinza de uma imagem em um sinal de comunicação possibilita o uso de técnicas de equalização cega de canais para a restauração de imagens. Neste trabalho, propõe-se o uso de um esquema para desconvolução cega de imagens baseado na combinação convexa de um equalizador cego com um equalizador no modo de decisão direta. A combinação também é adaptada de forma cega, o que possibilita o chaveamento automático entre os filtros componentes. Dessa forma, o esquema proposto é capaz de atingir o desempenho de um algoritmo de filtragem adaptativa supervisionada sem o conhecimento prévio da imagem original. O desempenho da combinação é ilustrado por meio de simulações, que comprovam a eficiência desse esquema quando comparado a outras soluções da literatura. / The goal of blind image deconvolution is to restore a degraded image without using information from the actual image or from the point spread function. The mapping of the gray levels of an image into a communication signal enables the use of blind equalization techniques for image restoration. In this work, we use a blind image deconvolution scheme based on the convex combination of a blind equalizer with an equalizer in the decision-directed mode. The combination is also blindly adapted, which enables automatic switching between the component filters. Thus, the proposed scheme is able to achieve the performance of a supervised adaptive filtering algorithm without prior knowledge of the original image. The performance of the combination is illustrated by simulations, which show the efficiency of this scheme when compared to other solutions in the literature.

Submitted by Rosivalda Pereira (mrs.pereira@ufma.br) on 2017-06-23T20:42:44Z No. of bitstreams: 1 CesarCasteloBranco.pdf: 11257769 bytes, checksum: 911c33f2f0ba5c1c0948888e713724f6 (MD5) / Made available in DSpace on 2017-06-23T20:42:44Z (GMT). No. of bitstreams: 1 CesarCasteloBranco.pdf: 11257769 bytes, checksum: 911c33f2f0ba5c1c0948888e713724f6 (MD5) Previous issue date: 2016-12-12 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPQ) / This work proposes new methodologies to optimize the choice of the parameters of the proportionate normalized least-mean-square (PNLMS) adaptive algorithms. The proposed approaches use procedures based on two optimization methods, namely, the golden section and tabu search methods. Such procedures are applied to determine the optimal parameters in each iteration of the adaptation process of the PNLMS and improved PNLMS (IPNLMS) algorithms. The objective function for the proposed procedures is based on the a posteriori estimation error. Performance studies carried out to evaluate the impact of the PNLMS and IPNLMS parameters in the behavior of these algorithms shows that, with the aid of optimization techniques to choose properly such parameters, the performance of these algorithms may be improved in terms of convergence speed for the identification of plants with high sparseness degree. The main goal of the proposed methodologies is to improve the distribution of the adaptation energy between the coefficients of the PNLMS and IPNLMS algorithms, using parameter values that lead to the minimal estimation error of each iteration of the adaptation process. Numerical tests performed (considering various scenarios in which the plant impulse response is sparse) show that the proposed methodologies achieve convergence speeds faster than the PNLMS and IPNLMS algorithms, and other algorithms of the PNLMS class, such as the sparseness controlled IPNLMS (SC-IPNLMS) algorithm. / Neste trabalho, novas metodologias para otimizar a escolha dos parâmetros dos algoritmos adaptativos LMS normalizados proporcionais (PNLMS) são propostas. As abordagens propostas usam procedimentos baseados em dois métodos de otimização, a saber, os métodos da razão áurea e da busca tabu. Tais procedimentos são empregados para determinar os parâmetros ótimos em cada iteração do processo de adaptação dos algoritmos PNLMS e PNLMS melhorado (IPNLMS). A função objetivo adotada pelos procedimentos propostos é baseada no erro de estimação a posteriori. O estudo de desempenho realizado para avaliar o impacto dos parâmetros dos algoritmos PNLMS e IPNLMS no comportamento dos mesmos mostram que, com o auxílio de técnicas de otimização para escolher adequadamente tais parâmetros, o desempenho destes algoritmos pode ser melhorado, em termos de velocidade de convergência, para a identificação de plantas com elevado grau de esparsidade. O principal objetivo das metodologias propostas é melhorar a distribuição da energia de ativação entre os coeficientes dos algoritmos PNLMS e IPNLMS, usando valores de parâmetros que levam ao erro de estimação mínimo em cada iteração do processo de adaptação. Testes numéricos realizados (considerando diversos cenários nos quais a resposta impulsiva da planta é esparsa) mostram que as metodologias propostas alcançam velocidades de convergência superiores às dos algoritmos PNLMS e IPNLMS, além de outros algoritmos da classe PNLMS, tais como o algoritmo IPNLMS com controle de esparsidade (SCIPNLMS).

The goal of this term project is create user friendly client of Twitter. They may use methods of machine learning as naive bayes classifier to mentions new interests tweets. For visualissation this tweets will be use hyperbolic trees and some others methods.

[pt] As redes da Internet das Coisas (IdC) incluem dispositivos inteligentes que contêm muitos sensores que permitem interagir com o mundo físico, coletando e processando dados de streaming em tempo real. O consumo total de energia e o custo desses sensores afetam o consumo de energia e o custo dos dispositivos IdC. O tipo de sensor determina a precisão da interface analógica e a resolução dos conversores analógico-digital (ADCs). A resolução dos ADCs tem um compromisso entre a precisão de inferência e o consumo de energia, uma vez que o consumo de energia dos ADCs depende do número de bits usados para representar amostras digitais. Nesta tese, apresentamos um esquema de aprendizado distribuído com eficiência energética usando sinais quantizados para redes da IdC. Em particular, desenvolvemos algoritmos de gradiente estocástico com reconhecimento de quantização distribuído (DQA-LMS) e de mínimos quadrados recursivos com reconhecimento de quantização distribuído (DQA-RLS) que podem aprender parâmetros de maneira eficiente em energia usando sinais quantizados com poucos bits, exigindo um baixo custo computacional. Além disso, desenvolvemos uma estratégia de compensação de viés para melhorar ainda mais o desempenho dos algoritmos propostos. Uma análise estatística dos algoritmos propostos juntamente com uma avaliação da complexidade computacional das técnicas propostas e existentes é realizada. Os resultados numéricos avaliam os algoritmos com reconhecimento de quantização distribuída em relação às técnicas existentes para uma tarefa de estimação de parâmetros em que os dispositivos IdC operam em um modo ponto a ponto. Também apresentamos um esquema de aprendizado federativo com eficiência energética usando sinais quantizados para redes de IdC. Desenvolvemos o algoritmo federated averaging LMS (QA-FedAvg-LMS) com reconhecimento de quantização para redes IdC estruturadas por configuração de aprendizado federativo em que os dispositivos IdC trocam suas estimativas com um servidor. Uma estratégia de compensação de viés para QA-FedAvg-LMS é proposta junto com sua análise estatística e a avaliação de desempenho em relação às técnicas existentes com resultados numéricos. / [en] Internet of Things (IoT) networks include smart devices that contain many sensors that allow them to interact with the physical world, collecting and processing streaming data in real time. The total energy-consumption and cost of these sensors affect the energy-consumption and the cost of IoT devices. The type of sensor determines the accuracy of the analog interface and the resolution of the analog-to-digital converters (ADCs). The ADC resolution requirement has a trade-off between sensing performance and energy consumption since the energy consumption of ADCs strongly depends on the number of bits used to represent digital samples. In this thesis, we present an energy-efficient distributed learning framework using coarsely quantized signals for IoT networks. In particular, we develop a distributed quantization-aware least-mean square (DQA-LMS) and a distributed quantization-aware recursive least-squares (DQA-RLS) algorithms that can learn parameters in an energy-efficient fashion using signals quantized with few bits while requiring a low computational cost. Moreover, we develop a bias compensation strategy to further improve the performance of the proposed algorithms. We then carry out a statistical analysis of the proposed algorithms along with a computational complexity evaluation of the proposed and existing techniques. Numerical results assess the distributed quantization-aware algorithms against existing techniques for distributed parameter estimation where IoT devices operate in a peer-to-peer mode. We also introduce an energy-efficient federated learning framework using coarsely quantized signals for IoT networks, where IoT devices exchange their estimates with a server. We then develop the quantization-aware federated averaging LMS (QA-FedAvg-LMS) algorithm to perform parameter estimation at the clients and servers. Furthermore, we devise a bias compensation strategy for QA-FedAvg-LMS, carry out its statistical analysis, and assess its performance against existing techniques with numerical results.

No description available.

Les émetteurs-récepteurs actuels tendent à devenir multi-standards c’est-àdireque plusieurs standards de communication peuvent cohabiter sur la même puce. Lespuces sont donc amenées à traiter des signaux de formes très différentes, et les composantsanalogiques subissent des contraintes de conception de plus en plus fortes associées au supportdes différentes normes. Les auto-interférences, c’est à dire les interférences généréespar le système lui-même, sont donc de plus en plus présentes, et de plus en plus problématiquesdans les architectures actuelles. Ces travaux s’inscrivent dans le paradigmede la « radio sale » qui consiste à accepter une pollution partielle du signal d’intérêtet à réaliser, par l’intermédiaire d’algorithmes, une atténuation de l’impact de ces pollutionsauto-générées. Dans ce manuscrit, on s’intéresse à différentes auto-interférences(phénomène de "spurs", de "Tx leakage", ...) dont on étudie les modèles numériques etpour lesquelles nous proposons des stratégies de compensation. Les algorithmes proposéssont des algorithmes de traitement du signal adaptatif qui peuvent être vus comme des« algorithmes de soustraction de bruit » basés sur des références plus ou moins précises.Nous dérivons analytiquement les performances transitionnelles et asymptotiques théoriquesdes algorithmes proposés. On se propose également d’ajouter à nos systèmes unesur-couche originale qui permet d’accélérer la convergence, tout en maintenant des performancesasymptotiques prédictibles et paramétrables. Nous validons enfin notre approchesur une puce dédiée aux communications cellulaires ainsi que sur une plateforme de radiologicielle. / Radio frequency transceivers are now massively multi-standards, which meansthat several communication standards can cohabit in the same environment. As a consequence,analog components have to face critical design constraints to match the differentstandards requirements and self-interferences that are directly introduced by the architectureitself are more and more present and detrimental. This work exploits the dirty RFparadigm : we accept the signal to be polluted by self-interferences and we develop digitalsignal processing algorithms to mitigate those aforementioned pollutions and improve signalquality. We study here different self-interferences and propose baseband models anddigital adaptive algorithms for which we derive closed form formulae of both transientand asymptotic performance. We also propose an original adaptive step-size overlay toimprove transient performance of our method. We finally validate our approach on a systemon chip dedicated to cellular communications and on a software defined radio.

Les émetteurs-récepteurs actuels tendent à devenir multi-standards c’est-àdireque plusieurs standards de communication peuvent cohabiter sur la même puce. Lespuces sont donc amenées à traiter des signaux de formes très différentes, et les composantsanalogiques subissent des contraintes de conception de plus en plus fortes associées au supportdes différentes normes. Les auto-interférences, c’est à dire les interférences généréespar le système lui-même, sont donc de plus en plus présentes, et de plus en plus problématiquesdans les architectures actuelles. Ces travaux s’inscrivent dans le paradigmede la « radio sale » qui consiste à accepter une pollution partielle du signal d’intérêtet à réaliser, par l’intermédiaire d’algorithmes, une atténuation de l’impact de ces pollutionsauto-générées. Dans ce manuscrit, on s’intéresse à différentes auto-interférences(phénomène de "spurs", de "Tx leakage", ...) dont on étudie les modèles numériques etpour lesquelles nous proposons des stratégies de compensation. Les algorithmes proposéssont des algorithmes de traitement du signal adaptatif qui peuvent être vus comme des« algorithmes de soustraction de bruit » basés sur des références plus ou moins précises.Nous dérivons analytiquement les performances transitionnelles et asymptotiques théoriquesdes algorithmes proposés. On se propose également d’ajouter à nos systèmes unesur-couche originale qui permet d’accélérer la convergence, tout en maintenant des performancesasymptotiques prédictibles et paramétrables. Nous validons enfin notre approchesur une puce dédiée aux communications cellulaires ainsi que sur une plateforme de radiologicielle. / Radio frequency transceivers are now massively multi-standards, which meansthat several communication standards can cohabit in the same environment. As a consequence,analog components have to face critical design constraints to match the differentstandards requirements and self-interferences that are directly introduced by the architectureitself are more and more present and detrimental. This work exploits the dirty RFparadigm : we accept the signal to be polluted by self-interferences and we develop digitalsignal processing algorithms to mitigate those aforementioned pollutions and improve signalquality. We study here different self-interferences and propose baseband models anddigital adaptive algorithms for which we derive closed form formulae of both transientand asymptotic performance. We also propose an original adaptive step-size overlay toimprove transient performance of our method. We finally validate our approach on a systemon chip dedicated to cellular communications and on a software defined radio.

Dans cette thèse nous développons des estimations d'erreur a posteriori par équilibrage de flux pour la poro-mécanique et la poro-plasticité.En se basant sur ces estimations, nous proposons des algorithmes adaptatifs pour la résolution numérique de problèmes en mécanique des sols.Le premier chapitre traite des problèmes en poro-élasticité linéaire.Nous obtenons une borne garantie sur l'erreur en utilisant des reconstructions équilibrées et $H({rm div})$-conformes de la vitesse de Darcy et du tenseur de contraintes mécaniques.Nous appliquons cette estimation dans un algorithme adaptif pour équilibrer les composantes de l'erreur provenant de la discrétisation en espace et en temps pour des simulations en deux dimensions.La contribution principale du chapitre porte sur la reconstruction symétrique du tenseur de contraintes.Dans le deuxième chapitre nous proposons une deuxième technique de reconstruction du tenseur de contraintes dans le cadre de l'élasticité nonlinéaire.En imposant la symétrie faiblement, cette technique améliore les temps de calcul et facilite l'implémentation.Nous démontrons l'éfficacité locale et globale des estimateurs obtenus avec cette reconstruction pour une grande classe de lois en hyperélasticité.En ajoutant un estimateur de l'erreur de linéarisation, nous introduisons des critères d'arrêt adaptatifs pour le solveur de linéarisation.Le troisième chapitre est consacré à l'application industrielle des résultats obtenus. Nous appliquons un algorithme adaptatif à des problèmes poro-mécaniques en trois dimensions avec des lois de comportement mécanique élasto-plastiques. / In this Ph.D. thesis we develop equilibrated flux a posteriori error estimates for poro-mechanical and poro-plasticity problems.Based on these estimations we propose adaptive algorithms for the numerical solution of problems in soil mechanics.The first chapter deals with linear poro-elasticity problems.Using equilibrated $H({rm div})$-conforming flux reconstructions of the Darcy velocity and the mechanical stress tensor, we obtain a guaranteed upper bound on the error.We apply this estimate in an adaptive algorithm balancing the space and time discretisation error components in simulations in two space dimensions.The main contribution of this chapter is the symmetric reconstruction of the stress tensor.In the second chapter we propose another reconstruction technique for the stress tensor, while considering nonlinear elasticity problems.By imposing the symmetry of the tensor only weakly, we reduce computation time and simplify the implementation.We prove that the estimate obtained using this stress reconstuction is locally and globally efficient for a wide range of hyperelasticity problems.We add a linearization error estimator, enabling us to introduce adaptive stopping criteria for the linearization solver.The third chapter adresses the industrial application of the obtained results.We apply an adaptive algorithm to three-dimensional poro-mechanical problems involving elasto-plastic mechanical behavior laws.

L’apprentissage adaptatif distribué sur les réseaux permet à un ensemble d’agents de résoudre des problèmes d’estimation de paramètres en ligne en se basant sur des calculs locaux et sur des échanges locaux avec les voisins immédiats. La littérature sur l’estimation distribuée considère essentiellement les problèmes à simple tâche, où les agents disposant de fonctions objectives séparables doivent converger vers un vecteur de paramètres commun. Cependant, dans de nombreuses applications nécessitant des modèles plus complexes et des algorithmes plus flexibles, les agents ont besoin d’estimer et de suivre plusieurs vecteurs de paramètres simultanément. Nous appelons ce type de réseau, où les agents doivent estimer plusieurs vecteurs de paramètres, réseau multitâche. Bien que les agents puissent avoir différentes tâches à résoudre, ils peuvent capitaliser sur le transfert inductif entre eux afin d’améliorer les performances de leurs estimés. Le but de cette thèse est de proposer et d’étudier de nouveaux algorithmes d’estimation distribuée sur les réseaux multitâches. Dans un premier temps, nous présentons l’algorithme diffusion LMS qui est une stratégie efficace pour résoudre les problèmes d’estimation à simple-tâche et nous étudions théoriquement ses performances lorsqu’il est mis en oeuvre dans un environnement multitâche et que les communications entre les noeuds sont bruitées. Ensuite, nous présentons une stratégie de clustering non-supervisé permettant de regrouper les noeuds réalisant une même tâche en clusters, et de restreindre les échanges d’information aux seuls noeuds d’un même cluster / Distributed adaptive learning allows a collection of interconnected agents to perform parameterestimation tasks from streaming data by relying solely on local computations and interactions with immediate neighbors. Most prior literature on distributed inference is concerned with single-task problems, where agents with separable objective functions need to agree on a common parameter vector. However, many network applications require more complex models and flexible algorithms than single-task implementations since their agents involve the need to estimate and track multiple objectives simultaneously. Networks of this kind, where agents need to infer multiple parameter vectors, are referred to as multitask networks. Although agents may generally have distinct though related tasks to perform, they may still be able to capitalize on inductive transfer between them to improve their estimation accuracy. This thesis is intended to bring forth advances on distributed inference over multitask networks. First, we present the well-known diffusion LMS strategies to solve single-task estimation problems and we assess their performance when they are run in multitask environments in the presence of noisy communication links. An improved strategy allowing the agents to adapt their cooperation to neighbors sharing the same objective is presented in order to attain improved learningand estimation over networks. Next, we consider the multitask diffusion LMS strategy which has been proposed to solve multitask estimation problems where the network is decomposed into clusters of agents seeking different