This thesis deals with Matlab application developed for simulation of the DVB-T channel coder and decoder. The first part of this thesis includes description of terrestrial digital video broadcasting system and comparison with analogue television. Channel coding and OFDM modulation, used in the DVB-T standard, is described in detail. Application developed in the Matlab environment is described in the second part. The application simulates data transfer of the DVB-T system. Results of the simulated transmission, using developed application are presented in the last part. Namely dependence of the BER on the S/N ratio, using various coder settings, was examined. Maximal possible data rate was determined for these various setting. All obtained values are graphically represented.

This paper concerns aliasing removal methods during the shadow displaying. Method of shadow mapping, its principles, procedure and mainly its drawbacks in the form of aliasing development are described. For the removal of this undesirable phenomenon, several aliasing suppressing methods are described - Percentage Closer Filter, Variance Shadow Map, Convulotion Shadow Map, Exponential Shadow Map a Bilateral Filter. I conclude my work with a proposal and implementation of a demonstrative application, which demonstrates the implemented results adequately. Also, the comparison of individual methods on the basis of their quality and computational demands is included.

Diploma thesis is aimed to trainable image segmentation using deep neural networks. In the paper is explained the principle of digital image processing and image segmentation. In the paper is also explained the principle of artificial neural network, model of artificial neuron, training and activation of artificial neural network. In practical part of the paper is created an algorithm of sliding window to generate sub-images from image from magnetic rezonance. Generated sub-images are used to train, test and validate of the model of neural network. In practical part of the paper si created the model of the artificial neural network, which is used to trainable image segmentation. Model of the neural network is created using the Deeplearning4j library and it is optimized to parallel training using Spark library.

This work is concerned with the applicability of several approximation methods (finite section method, Galerkin and collocation methods with maximum defect splines for uniform and non uniform meshes) to operators belonging to the closed subalgebra generated by operators of multiplication bz piecewise continuous functions and convolution operators also with piecewise continuous generating function.

This dissertation is concerned with invertibility and one-sided invertibility of Toeplitz operators over the half plane whose generating functions admit homogenous discontinuities, and with stability of their pseudo finite sections. The invertibility criterium is given in terms of invertibility of a family of one dimensional Toeplitz operators with piecewise continuous generating functions. The one-sided invertibility criterium is given it terms of constraints on the partial indices of certain Toeplitz operator valued function.

Lineare hyperbolische partielle Differentialgleichungen in homogenen Medien, beispielsweise die Wellengleichung, die die Ausbreitung und die Streuung akustischer Wellen beschreibt, können im Zeitbereich mit Hilfe von Randintegralgleichungen formuliert werden. Im ersten Hauptteil dieser Arbeit stellen wir eine effiziente Möglichkeit vor, numerische Approximationen solcher Gleichungen zu implementieren, wenn das Huygens-Prinzip nicht gilt. Wir nutzen die Faltungsquadraturmethode für die Zeitdiskretisierung und eine Galerkin-Randelement-Methode für die Raumdiskretisierung. Mit der Faltungsquadraturmethode geht eine diskrete Faltung der Faltungsgewichte mit der Randdichte einher. Bei Gültigkeit des Huygens-Prinzips konvergieren die Gewichte exponentiell gegen null, sofern der Index hinreichend groß ist. Im gegenteiligen Fall, das heißt bei geraden Raumdimensionen oder wenn Dämpfungseffekte auftreten, kann kein Verschwinden der Gewichte beobachtet werden. Das führt zu Schwierigkeiten bei der effizienten numerischen Behandlung. Im ersten Hauptteil dieser Arbeit zeigen wir, dass die Kerne der Faltungsgewichte in gewisser Weise die Fundamentallösung im Zeitbereich approximieren und dass dies auch zutrifft, wenn beide bezüglich der räumlichen Variablen abgeleitet werden. Da die Fundamentallösung zudem für genügend große Zeiten, etwa nachdem die Wellenfront vorbeigezogen ist, glatt ist, schließen wir Gleiches auch in Bezug auf die Faltungsgewichte, die wir folglich mit hoher Genauigkeit und wenigen Interpolationspunkten interpolieren können. Darüber hinaus weisen wir darauf hin, dass zur weiteren Einsparung von Speicherkapazitäten, insbesondere bei Langzeitexperimenten, der von Schädle et al. entwickelte schnelle Faltungsalgorithmus eingesetzt werden kann. Wir diskutieren eine effiziente Implementierung des Problems und zeigen Ergebnisse eines numerischen Langzeitexperimentes. Im zweiten Hauptteil dieser Arbeit beschäftigen wir uns mit Transmissionsproblemen der Wellengleichung im Freiraum. Solche Probleme werden gewöhnlich derart behandelt, dass der Freiraum, wenn nötig durch Einführen eines künstlichen Randes, in ein unbeschränktes Außengebiet und ein beschränktes Innengebiet geteilt wird mit dem Ziel, eventuelle Inhomogenitäten oder Nichtlinearitäten des Materials vollständig im Innengebiet zu konzentrieren. Wir werden eine Lösungsstrategie vorstellen, die es erlaubt, die aus der Teilung resultierenden Teilprobleme so weit wie möglich unabhängig voneinander zu behandeln. Die Kopplung der Teilprobleme erfolgt über Transmissionsbedingungen, die auf dem ihnen gemeinsamen Rand vorgegeben sind. Wir diskutieren ein Kopplungsverfahren, das auf verschiedene Diskretisierungsschemata für das Innen- und das Außengebiet zurückgreift. Wir werden insbesondere ein explizites Verfahren im Innengebiet einsetzen, im Gegensatz zum Außengebiet, bei dem wir ein auf ein Mehrschrittverfahren beruhendes Faltungsquadraturverfahren nutzen. Die Kopplung erfolgt nach der Strategie von Johnson und Nédélec, bei der die direkte Randintegralmethode zum Einsatz kommt. Diese Strategie führt auf ein unsymmetrische System. Wir analysieren das diskrete Problem hinsichtlich Stabilität und Konvergenz und unterstreichen die Einsatzfähigkeit des Kopplungsalgorithmus mit der Durchführung numerischer Experimente.

Le sujet de la thèse est l'extraction et la segmentation des vêtements à partir d'images en utilisant des techniques de la vision par ordinateur, de l'apprentissage par ordinateur et de la description d'image, pour la recommandation de manière non intrusive aux utilisateurs des produits similaires provenant d'une base de données de vente. Nous proposons tout d'abord un extracteur d'objets dédié à la segmentation de la robe en combinant les informations locales avec un apprentissage préalable. Un détecteur de personne localises des sites dans l'image qui est probable de contenir l'objet. Ensuite, un processus d'apprentissage intra-image en deux étapes est est développé pour séparer les pixels de l'objet de fond. L'objet est finalement segmenté en utilisant un algorithme de contour actif qui prend en compte la segmentation précédente et injecte des connaissances spécifiques sur la courbure locale dans la fonction énergie. Nous proposons ensuite un nouveau framework pour l'extraction des vêtements généraux en utilisant une procédure d'ajustement globale et locale à trois étapes. Un ensemble de modèles initialises un processus d'extraction d'objet par un alignement global du modèle, suivi d'une recherche locale en minimisant une mesure de l'inadéquation par rapport aux limites potentielles dans le voisinage. Les résultats fournis par chaque modèle sont agrégés, mesuré par un critère d'ajustement globale, pour choisir la segmentation finale. Dans notre dernier travail, nous étendons la sortie d'un réseau de neurones Fully Convolutional Network pour inférer le contexte à partir d'unités locales (superpixels). Pour ce faire, nous optimisons une fonction énergie, qui combine la structure à grande échelle de l'image avec le local structure superpixels, en recherchant dans l'espace de toutes les possibilité d'étiquetage. De plus, nous introduisons une nouvelle base de données RichPicture, constituée de 1000 images pour l'extraction de vêtements à partir d'images de mode. Les méthodes sont validées sur la base de données publiques et se comparent favorablement aux autres méthodes selon toutes les mesures de performance considérées. / The topic of the thesis is the extraction and segmentation of clothing items from still images using techniques from computer vision, machine learning and image description, in view of suggesting non intrusively to the users similar items from a database of retail products. We firstly propose a dedicated object extractor for dress segmentation by combining local information with a prior learning. A person detector is applied to localize sites in the image that are likely to contain the object. Then, an intra-image two-stage learning process is developed to roughly separate foreground pixels from the background. Finally, the object is finely segmented by employing an active contour algorithm that takes into account the previous segmentation and injects specific knowledge about local curvature in the energy function.We then propose a new framework for extracting general deformable clothing items by using a three stage global-local fitting procedure. A set of template initiates an object extraction process by a global alignment of the model, followed by a local search minimizing a measure of the misfit with respect to the potential boundaries in the neighborhood. The results provided by each template are aggregated, with a global fitting criterion, to obtain the final segmentation.In our latest work, we extend the output of a Fully Convolution Neural Network to infer context from local units(superpixels). To achieve this we optimize an energy function,that combines the large scale structure of the image with the locallow-level visual descriptions of superpixels, over the space of all possiblepixel labellings. In addition, we introduce a novel dataset called RichPicture, consisting of 1000 images for clothing extraction from fashion images.The methods are validated on the public database and compares favorably to the other methods according to all the performance measures considered.

Indiana University-Purdue University Indianapolis (IUPUI) / In recent years, deep learning models have become popular in the real-time embedded application, but there are many complexities for hardware deployment because of limited resources such as memory, computational power, and energy. Recent research in the field of deep learning focuses on reducing the model size of the Convolution Neural Network (CNN) by various compression techniques like Architectural compression, Pruning, Quantization, and Encoding (e.g., Huffman encoding). Network pruning is one of the promising technique to solve these problems. This thesis proposes methods to prune the convolution neural network (SqueezeNet) without introducing network sparsity in the pruned model. This thesis proposes three methods to prune the CNN to decrease the model size of CNN without a significant drop in the accuracy of the model. 1: Pruning based on Taylor expansion of change in cost function Delta C. 2: Pruning based on L2 normalization of activation maps. 3: Pruning based on a combination of method 1 and method 2. The proposed methods use various ranking methods to rank the convolution kernels and prune the lower ranked filters afterwards SqueezeNet model is fine-tuned by backpropagation. Transfer learning technique is used to train the SqueezeNet on the CIFAR-10 dataset. Results show that the proposed approach reduces the SqueezeNet model by 72% without a significant drop in the accuracy of the model (optimal pruning efficiency result). Results also show that Pruning based on a combination of Taylor expansion of the cost function and L2 normalization of activation maps achieves better pruning efficiency compared to other individual pruning criteria and most of the pruned kernels are from mid and high-level layers. The Pruned model is deployed on BlueBox 2.0 using RTMaps software and model performance was evaluated.

Equatorial plasma bubbles (EPBs) are large structures consisting of depleted plasma that generally form on the nightside of Earth's ionosphere along magnetic field lines in the upper thermosphere/ionosphere. While referred to as `bubbles', EPBs tend to be longer along magnetic latitudes and narrower along magnetic longitudes which are on the order of thousands and hundreds of kilometers, respectively. EPBs are a well documented occurrence with observations spanning many decades. As such, much is known about their general behavior, seasonal variation of occurrences, increasing/decreasing occurrences with increasing/decreasing solar activity, and their ability to interact and interfere with radio waves such as GPS. This dissertation expands on this understanding by focusing on the detection and tracking of EPBs in the upper thermosphere/ionosphere along equatorial to low latitudes. To do this, far ultraviolet (FUV) emission observations of the recombination of O$^+$ with electrons via the Global-Scale Observations of the Limb and Disk (GOLD) mission are analyzed. GOLD provides consistent data from geostationary orbit with the eastern region of the Americas, Atlantic, and western Africa. The optical data can be used to pick out gradients in brightness along the 135.6 nm wavelength which correlate with the location of EPBs in the nightside ionosphere. The dissertation provides a novel method to look at and analyze 2-dimensional data with inconsistent time-steps for EPB detection and tracking. During development, preprocessing of large scale (multiple years) data proved to be the largest time sync. To that end, this dissertation tests the possibility of using convolution neural networks for detection of EPBs with the end goal of reducing the amount of preprocessing necessary. Further, data from the Ionospheric Connection Explorer's (ICON's) ion velocity meter (IVM) are compared to EPBs detected via GOLD to understand how the ambient plasma around the EPBs behave. Along with the ambient plasma, zonal and meridional thermospheric winds observed by ICON's Michelson Interferometer for Global High-resolution Thermospheric Imaging (MIGHTI) instrument are analyzed in conjunction with the same EPBs to understand how winds coincident with EPBs behave. An analysis of winds before EPBs form is also done to observe the potential for both zonal and meridional winds' ability to suppress and amplify EPB formation. / Doctor of Philosophy / Equatorial plasma bubbles (EPBs) are large structures that generally form during post- sunset along Earth's magnetic equator. While referred to as `bubbles', EPBs tend to be thousands of kilometers from north to south and hundreds of kilometers from east to west and well over a thousands kilometers in altitude. EPBs are a well documented occurrence with observations spanning many decades. This includes their ability to interfere with radar and GPS. This dissertation expands on the scientific community's understanding by focusing on the detection and tracking of EPBs along the magnetic equator. To do this, observations from the NASA Global-Scale Observations of the Limb and Disk (GOLD) mission are analyzed. GOLD provides consistent observations looking over the eastern region of the Americas, Atlantic, and western Africa. A unique method to look at and analyze this data for EPB detection and tracking is developed. This dissertation also tests the possibility of using machine learning for detection of EPBs. Further, data from the NASA Ionospheric Connection Explorer (ICON) mission is compared to EPBs detected via GOLD to understand how the behavior of the upper atmosphere and the conductive region therein, known as the ionosphere, interact with the EBPs themselves.

Dans cette thèse, nous nous intéressons au problème d'estimation de densité dans le modèle de convolution. Ce cadre correspond aux modèles avec erreurs de mesures additives, c'est-à-dire que nous observons une version bruitée de la variable d'intérêt. Pour mener notre étude, nous adoptons le point de vue de l'estimation non-paramétrique adaptative qui repose sur des procédures de sélection de modèle développées par Birgé & Massart ou sur les méthodes de Lepski. Cette thèse se divise en deux parties. La première développe des méthodes spécifiques d'estimation adaptative quand les variables d'intérêt et les erreurs sont des variables aléatoires positives. Ainsi nous proposons des estimateurs adaptatifs de la densité ou encore de la fonction de survie dans ce modèle, puis de fonctionnelles linéaires de la densité cible. Enfin nous suggérons une procédure d'agrégation linéaire. La deuxième partie traite de l'estimation adaptative de densité dans le modèle de convolution lorsque la loi des erreurs est inconnue. Dans ce cadre il est supposé qu'un échantillon préliminaire du bruit est disponible ou que les observations sont disponibles sous forme de données répétées. Les résultats obtenus pour des données répétées dans le modèle de convolution permettent d'élargir cette méthodologie au cadre des modèles linéaires mixtes. Enfin cette méthode est encore appliquée à l'estimation de la densité de somme de variables aléatoires observées avec du bruit. / In this thesis, we are interested in nonparametric adaptive estimation problems of density in the convolution model. This framework matches additive measurement error models, which means we observe a noisy version of the random variable of interest. To carry out our study, we follow the paradigm of model selection developped by Birgé & Massart or criterion based on Lepski's method. The thesis is divided into two parts. In the first one, the main goal is to build adaptive estimators in the convolution model when both random variables of interest and errors are distributed on the nonnegative real line. Thus we propose adaptive estimators of the density along with the survival function, then of linear functionals of the target density. This part ends with a linear density aggregation procedure. The second part of the thesis deals with adaptive estimation of density in the convolution model when the distribution is unknown and distributed on the real line. To make this problem identifiable, we assume we have at hand either a preliminary sample of the noise or we observe repeated data. So, we can derive adaptive estimation with mild assumptions on the noise distribution. This methodology is then applied to linear mixed models and to the problem of density estimation of the sum of random variables when the latter are observed with an additive noise.