This thesis is in the field of computer vision, focusing on the problems of optical flow estimation. Optical flow is a notoriously difficult 2D problem since it’s inherently underconstrained. To introduce the concept of cyclopean optical flow, we will downgrade the 2D into 1D to make it more accessible. It proposes a new approach based on a "cyclopean" frame of reference. We apply a constrained gradient-based technique to solve 1D optical flow, for which the constraints are gradient behavior and correlation score. This thesis focuses on the fundamental problem of ensuring that the gradient remains usable in an interval large enough to cover the spatial displacement of motion. The proposed "cyclopean" approach does not enforce optical measurements over a fixed grid, which results in more reliable results. To further increase the allowed motion interval, we propose a pyramidal constraint that allows solving over a coarse-to-fine approach. We solved over aerial imagery, Sintel data-set, and Sintel data-set when artificially displaced 10% of the ground truth. This work is developed in the "continuous" framework commonly used for small motion optical flow. Our results showed good management of false positives while maintaining a good amount of convergence density. However, our method isn’t as precise as the current state-of-the-art benchmarks, as it specializes in very small motions. Also, it’s important to mention versatility comes with the concept of "continuous" representation. This allows us to select regions to be solved, opening the possibility of adapting to the spectrum of sparse or dense optical flow. From this study point of view, we can highlight traditional methods have relevance even in the deep learning era, offering a new set of tools to exploit on the pursue of solving optical flow. / Ce mémoire s’intéresse au domaine de la vision par ordinateur, et plus particulièrement l’estimation du flux optique. Le flux optique est un problème 2D notoirement difficile, car il est intrinsèquement sous-contraint. Pour introduire la notion de flux optique cyclopéen, nous allons considérer le problème en 1D pour éliminer le problème d’ouverture lié au mouvement 2D. Nous proposons une nouvelle approche basée sur un référentiel « cyclopéen », basée sur gradient calculé dans un espace continu pour résoudre le flux optique 1D. Ce mémoire se concentre a garantir que le gradient reste utilisable dans un intervalle suffisamment grand pour couvrir le déplacement spatial du mouvement. Lors de la résolution sur une approche coarse-to-fine, une représentation pyramidale est utilisée. Les résultats sur des images aériennes ainsi que des données synthétiques sont prometteurs. Ce travail se distingue des tendances actuelles en flux optique parle fait qu’il se spécialise pour les flux optiques à faible mouvement. Nos résultats ont montré une bonne gestion des faux positifs tout en conservant une bonne densité. Nous considérons que la fiabilité des mesures de mouvement est très élevée, ce qui est au moins aussi important que la précision elle-même dans beaucoup d’applications. Ainsi, la polyvalence de la représentation "continue" permet de mieux contrôler la densité obtenue en fonction de la scène analysée. À notre avis, cette approche, qui complète les méthodes traditionnelles, ouvrira la voie à de nouvelles approches en apprentissage profond.

Le Syndrome de l’X Fragile (SXF) est la forme héréditaire de retard mental la plus fréquente et la cause la mieux caractérisée de troubles du spectre autistique (TSA). Elle est causée par une mutation causant l’inactivation du gène Fmr1 (codant pour la protéine FMRP). La sensibilité accrue aux stimuli sensoriels est une caractéristique importante du SXF et des TSA, mais les mécanismes sous-jacents sont encore mal compris. Nous avons constaté que la suppression du gène Fmr1 entrainait une hyperexcitabilité sensorielle dans le modèle murin du SXF. Les souris Fmr1KO nécessitaient significativement moins d'informations tactiles pour l'exploration haptique, et les représentations évoquées par les informations tactiles provenant des vibrisses dans le cortex somatosensoriel primaire (S1) se propageaient à une vitesse plus élevée chez les souris Fmr1KO par rapport aux souris témoins sauvages.Au niveau cellulaire, il a été montré que les ARNm de plusieurs sous-unités de canaux ioniques (par exemple HCN1, KCNMA1) jouant un rôle clé dans le traitement de l'information dendritique / neuronale étaient des cibles de la protéine FMRP (Liao et al, 2008; Darnell et al, 2011). Sur la base de ces observations, nous avons étudié les canalopathies comme une caractéristique importante du SXF. Nous avons testé de possibles dysfonctionnement des canaux ioniques, et leurs conséquences sur le traitement de l'information dendritique dans les neurones pyramidaux du néocortex de la couche 5 chez les souris Fmr1KO, en utilisant une combinaison d’approches électrophysiologiques et d’imagerie calcique bi-photonique. Nos résultats ont montré que les dendrites des neurones pyramidaux du S1 étaient hyperexcitables, facilitant ainsi le couplage des entrées d’information synaptique à la génération de potentiel d'action en sortie dans les neurones. Cette altération était, au moins en partie, attribuable à un dysfonctionnement des canaux Ih et BKCa et a été partiellement restaurée par l'activation pharmacologique des canaux BKCa. Ces résultats plaident en faveur d'un rôle nouveau et crucial des canalopathies dans l'expression de l'hyperexcitabilité sensorielle dans le SXF. / Fragile X Syndrome (FXS) is the most common form of inherited mental retardation syndrome and most well characterized cause of Autism Spectrum Disorders (ASD), and it is caused by a silencing mutation of the gene Fmr1 (encoding the protein FMRP). Increased sensitivity to sensory stimuli is a prominent feature of FXS and ASD, but its underlying mechanisms are poorly understood. We found that deletion of the Fmr1 gene results in somatosensory hyper-excitability in a mouse model for FXS. Fmr1 knockout (Fmr1KO) mice required significantly less tactile information for haptic exploration, and touch-evoked whisker representations in the primary somatosensory cortex (S1) spread with increased velocity in Fmr1KO mice compared to wild-type control. At the cellular level, it has been shown that the mRNAs of several ion channel subunits (e.g. HCN1, KCNMA1) playing key roles in dendritic/neuronal information processing are regulated by FMRP (Liao et al., 2008; Darnell et al., 2011). Based on these observations, we investigated channelopathies as a prominent feature of FXS. We probed ion channel dysfunction, and its consequence for dendritic information processing in neocortical pyramidal neurons of layer 5 in Fmr1KO mice, using a combination of electrophysiological and 2-photon calcium imaging approaches. Our results showed that dendrites of S1 pyramidal neurons were hyper-excitable, facilitating the coupling of synaptic input to the generation of action potential output in these neurons. This defect was, at least in part, attributable to a dysfunction of Ih channels and BKCa channels and was partially rescued by pharmacological activation of BKCa channels. These findings argue for a novel and critical role for channelopathies in the expression of sensory hyper-excitability in FXS.

Weather radar imagery is important for several remote sensing applications including tracking of storm fronts and radar echo classification. In particular, tracking of precipitation events is useful for both forecasting and classification of rain/non-rain events since non-rain events usually appear to be static compared to rain events. Recent weather radar imaging-based forecasting approaches [3] consider that precipitation events can be modeled as a combination of localized functions using Radial Basis Function Neural Networks (RBFNNs). Tracking of rain events can be performed by tracking the parameters of these localized functions. The RBFNN-based techniques used in forecasting are not only computationally expensive, but also moderately effective in modeling small size precipitation events. In this thesis, an existing RBFNN technique [3] was implemented to verify its computational efficiency and forecasting effectiveness. The feasibility of modeling precipitation events using RBFNN effectively was evaluated, and several modifications to the existing technique have been proposed.

The foundation stone of this thesis is to perform a financial analysis of Czech non-financial companies for the years from 2007 to 2009 in order to evaluate development of the basic groups of financial analysis ratios before and after the outbreak of the financial crisis. Based on this analysis, I analyze the changes of the selected ratios in relation with the macroeconomic development. This is carried out by means of the description in words and for some significant changes using correlation analysis with GDP. The analysis results in the findings that the changes of ratios of return on assets and return on equity were negative in the analyzed period and for each year the factors of these changes are tracked using the pyramidal decomposition. In the statistical part of my thesis I partially succeeded in proving linear correlation between return on assets and GDP and a lower confidence level linear correlation between personnel cost to revenues ratio and GDP. Also, there were some correlations between individual ratios found to be significant.

No description available.

Le traitement de l'information sensorielle par le cortex cérébral requiert l'activation harmonieuse de micro-circuits neuronaux excitateurs et inhibiteurs interconnectés, ciblant les neurones pyramidaux de couche 5. Ces derniers élaborent les signaux de sortie corticaux et reçoivent un ratio de 20% d'excitation (E) et 80% d'inhibition (I). La dérégulation de cette balance E-I ou du système sérotoninergique conduit à des neuropathologies telles la dépression et la schizophrénie mais les interrelations entre la sérotonine et la balance E-I sont inconnues. Nous avons montré que la 5-HT endogène module la balance E-I en fonction du type de récepteur 5-HT recruté (1A, 2A, 3, 4, 7) et de sa localisation spécifique dans la colonne corticale. Ces données électrophysiologiques constituent la première évidence pour une action modulatrice fine de la sérotonine corticale sur la balance E-I et révèle la ségrégation fonctionnelle des récepteurs 5-HT dans les réseaux de neurones sensoriels.

Electronic imaging technologies are faced with the challenge of power consumption when transmitting large amounts of image data from the acquisition imager to the display or processing devices. This is especially a concern for portable applications, and becomes more prominent in increasingly high-resolution, high-frame rate imagers. Therefore, new sampling techniques are needed to minimize transmitted data, while maximizing the conveyed image information. <br /><br /> From this point of view, two approaches have been proposed and implemented in this thesis: <ol> <li> A system-level approach, in which the classical 1D row sampling CMOS imager is modified to a 2D ring sampling pyramidal architecture, using the same standard three transistor (3T) active pixel sensor (APS). </li> <li> A device-level approach, in which the classical orthogonal architecture has been preserved while altering the APS device structure, to design an expandable multiresolution image sensor. </li> </ol> A new scanning scheme has been suggested for the pyramidal image sensor, resulting in an intrascene foveated dynamic range (FDR) similar in profile to that of the human eye. In this scheme, the inner rings of the imager have a higher dynamic range than the outer rings. The pyramidal imager transmits the sampled image through 8 parallel output channels, allowing higher frame rates. The human eye is known to have less sensitivity to oblique contrast. Using this fact on the typical oblique distribution of fixed pattern noise, we demonstrate lower perception of this noise than the orthogonal FPN distribution of classical CMOS imagers. <br /><br /> The multiresolution image sensor principle is based on averaging regions of low interest from frame-sampled image kernels. One pixel is read from each kernel while keeping pixels in the region of interest at their high resolution. This significantly reduces the transferred data and increases the frame rate. Such architecture allows for programmability and expandability of multiresolution imaging applications.

Electronic imaging technologies are faced with the challenge of power consumption when transmitting large amounts of image data from the acquisition imager to the display or processing devices. This is especially a concern for portable applications, and becomes more prominent in increasingly high-resolution, high-frame rate imagers. Therefore, new sampling techniques are needed to minimize transmitted data, while maximizing the conveyed image information. <br /><br /> From this point of view, two approaches have been proposed and implemented in this thesis: <ol> <li> A system-level approach, in which the classical 1D row sampling CMOS imager is modified to a 2D ring sampling pyramidal architecture, using the same standard three transistor (3T) active pixel sensor (APS). </li> <li> A device-level approach, in which the classical orthogonal architecture has been preserved while altering the APS device structure, to design an expandable multiresolution image sensor. </li> </ol> A new scanning scheme has been suggested for the pyramidal image sensor, resulting in an intrascene foveated dynamic range (FDR) similar in profile to that of the human eye. In this scheme, the inner rings of the imager have a higher dynamic range than the outer rings. The pyramidal imager transmits the sampled image through 8 parallel output channels, allowing higher frame rates. The human eye is known to have less sensitivity to oblique contrast. Using this fact on the typical oblique distribution of fixed pattern noise, we demonstrate lower perception of this noise than the orthogonal FPN distribution of classical CMOS imagers. <br /><br /> The multiresolution image sensor principle is based on averaging regions of low interest from frame-sampled image kernels. One pixel is read from each kernel while keeping pixels in the region of interest at their high resolution. This significantly reduces the transferred data and increases the frame rate. Such architecture allows for programmability and expandability of multiresolution imaging applications.

Epilepsy is characterized by the hyperexcitability of individual neurons and hyper synchronization of groups of neurons (networks). The acquired changes that take place at molecular, cellular and network levels are important for the induction and maintenance of epileptic activity in the brain. Epileptic activity is known to alter the intrinsic properties and signaling of neurons. Understanding acquired changes that cause epilepsy may lead to innovative strategies to prevent or cure this neurological disorder. Advances in in vitro electrophysiological techniques together with experimental models of epilepsy are indispensible tools to understand molecular, cellular and network mechanisms that underlie epileptiform activity. The aim of the study was to investigate the epileptiform activity induced alterations in Ca2+ dynamics in apical dendrites of hippocampal subicular pyramidal neurons in slices and changes in network properties of cultured hippocampal neurons. We have also made attempts to develop an in vitro model of epilepsy using organotypic hippocampal slice cultures. In the first part of the present study, investigations on the basic properties of dendritic Ca2+ signaling in subicular pyramidal neurons during epileptiform activity are described. Subiculum, a part of the hippocampal formation is present, adjacent to the CA1 subfield. It acts as a transition zone between the hippocampus and entorhinal cortex. It receives inputs directly from the CA1 region, the entorhinal cortex, subcortical and other cortical areas. Several forms of evidences support the role of subiculum in temporal lobe epilepsy. Pronounced neuronal loss has been reported in various regions of the hippocampal formation (CA1 and CA3) leaving the subiculum generally intact in human epileptic tissue. It has been observed that epileptic activity is generated in subiculum in cases where the CA3 and CA1 regions are damaged or even absent. However, it is not clear how subicular neurons protect themselves from epileptic activity induced neuronal death. It is widely accepted that epileptiform activity induced neuronal damage is a result of an abnormally large influx of Ca2+ into neuronal compartments. In the present study, combined hippocampus / entorhinal cortical brain slices were exposed to zero Mg2+ + 4-amino pyridine artificial cerebrospinal fluid (ACSF) to generate spontaneous epileptiform discharges. Whole cell current-clamp recordings combined with Ca2+ imaging experiments (by incorporating Oregon green BAPTA-1 in the recording pipette) were performed on subicular pyramidal neurons to understand the changes in [Ca2+]i transients elicited in apical dendrites, in response to spontaneous epileptic discharges. To understand the changes occurring with respect to control, experiments were performed (in both control and in vitro epileptic conditions) where [Ca2+]i transients in dendrites were elicited by back propagating action potentials following somatic current injections. The results show clear distance-dependent changes in decay kinetics of [Ca2+]i transients (τdecay), without change in the amplitude of the [Ca2+]i transients, in distal parts (95–110 µm) compared to proximal segments (30–45 µm) of apical dendrites of subicular pyramidal neurons under in vitro epileptic condition, but not in control conditions. Pharmacological agents that block Ca2+ transporters viz. Na+/Ca2+ exchangers (Benzamil), plasma membrane Ca2+-ATPase pumps (Calmidazolium) and smooth endoplasmic reticulum Ca2+-ATPase pumps (Thapsigargin) were applied locally to the proximal and distal part of the apical dendrites in both experimental conditions to understand the molecular aspects of the Ca2+ extrusion mechanisms. The relative contribution of Na+/Ca2+ exchangers in Ca2+ extrusion was higher in the distal apical dendrite in in vitro epileptic condition. Using computer simulations with NEURON, biophysically realistic models were built to understand how faster decay of [Ca2+]i transients in the distal part of apical dendrite associated with [Ca2+]i extrusion mechanisms affect excitability of the neurons. With a linear increase in the density of Na+/Ca2+ exchangers along the apical dendrite, the decrease in τ decay values of [Ca2+]i transients in distal regions seen in experimental epileptic condition was reproduced in simulation. This linear increase in Na+/Ca2+ exchangers lowered the threshold for firing in response to consecutive synaptic inputs to the distal apical dendrite. Our results thus, show the existence of a novel neuroprotective mechanism in distal parts of the apical dendrite of subicular pyramidal neurons under in vitro epileptic condition with the Na+/Ca2+ exchangers being the major contributors to this mechanism. Although the enhanced contribution of Na+/Ca2+ exchangers helps the neuron in removing excess [Ca2+]i loads, it paradoxically makes the neuron hyperexcitable to synaptic inputs in the distal parts of the apical dendrites. Thus, the Na+/Ca2+ exchangers may actually protect subicular pyramidal neurons and at the same time contribute to the maintenance of epileptiform activity. In the second part of the study, neuronal network topologies and connectivity patterns were explored in control and glutamate injury induced epileptogenic hippocampal neuronal networks, cultured on planar multielectrode array (8×8) probes. Hyper synchronization of neuronal networks is the hallmark of epilepsy. To understand hyper synchronization and connectivity patterns of neuronal networks, electrical activity from multiple neurons were monitored simultaneously. The electrical activity recorded from a single electrode mainly consisted of randomly fired single spikes and bursts of spikes. Simultaneous measurement of electrical activity from all the 64 electrodes revealed network bursts. A network burst represents the period (lasting for 0.1–0.2 s) of synchronized activity in the network and, during this transient period, maximum numbers of neurons interact with each other. The network bursts were observed in both control and in vitro epileptic networks, but the frequency of network bursts was more in the latter, compared to former condition. Time stamps of individual spikes (from all 64 electrodes) during such time-aligned network burst were collected and stored in a matrix and used to construct the network topology. Connectivity maps were obtained by analyzing the spike trains using cross-covariance analysis and graph theory methods. Analysis of degree distribution, which is a measure of direct connections between electrodes in a neuronal network, showed exponential and Gaussian distributions in control and in vitro epileptic networks, respectively. Quantification of number of direct connections per electrode revealed that the in vitro epileptic networks showed much higher number of direct connections per electrode compared to control networks. Our results suggest that functional two-dimensional neuronal networks in vitro are not scale-free (not a power law degree distribution). After brief exposure to glutamate, normal hippocampal neuronal networks became hyperexcitable and fired a larger number of network bursts with altered network topology. Quantification of clustering coefficient and path length in these two types of networks revealed that the small-world network property was lost once the networks become epileptic and this was accompanied by a change from an exponential to a Gaussian network. In the last part of the study, we have explored if an excitotoxic glutamate injury (20 µM for 10 min) that produces spontaneous, recurrent, epileptiform discharges in cultured hippocampal neurons can induce epileptogenesis in hippocampal neurons of organotypic brain slice cultures. In vitro models of epilepsy are necessary to understand the mechanisms underlying seizures, the changes in brain structure and function that underlie epilepsy and are the best methods for developing new antiseizure and antiepileptogenic strategies. Glutamate receptor over-activation has been strongly associated with epileptogenesis. Recent studies have shown that brief exposure of dissociated hippocampal neurons in culture to glutamate (20 µM for 10 min) induces epileptogenesis in surviving neurons. Our aim was to extend the in vitro model of glutamate injury induced epilepsy to the slice preparations with intact brain circuits. Patch clamp technique in current-clamp mode was employed to monitor the expression of spontaneous epileptiform discharges from CA1 and CA3 neurons using several combinations of glutamate injury protocols. The results presented here represent preliminary efforts to standardize the glutamate injury protocol for inducing epileptogenesis in organotypic slice preparations. Our results indicate that glutamate injury protocols that induced epileptogenesis in dissociated hippocampal neurons in culture failed to turn CA1 and CA3 neurons of organotypic brain slice cultures epileptic. We also found that the CA1 and CA3 neurons of organotypic brain slice cultures are resilient to induction of epileptogenesis by glutamate injury protocols with 10 times higher concentrations of glutamate (200µM) than that used for neuronal cultures and long exposure periods (upto 30 min). These results clearly show that the factors involved in induction of epileptiform activity after glutamate injury in neuronal cultures and those involved in making the neurons in organotypic slices resilient to such insults are different, and understanding them could give vital clues about epileptogenesis and its control. The resilience of CA1 and CA3 neurons seen could be due to differences in homeostatic plasticity that operate in both these experimental systems. However, further studies are required to corroborate this hypothesis.

Lokale Pyramidenzellen sind die Hauptneurone des Hippocampus und können durch ihre Position und die Morphologie ihrer Dendriten als CA1 und CA3 Pyramidenzellen identifiziert werden. Die Dendriten der exzitatorischen Pyramidenzellen sind mit postsynaptischen Vorwölbungen, den so genannten Spines, bedeckt, welche in einem spezifischen Verteilungsmuster angeordnet sind. Neurotoxine wie das C3 Toxin von Clostridium botulinum sind funktionelle Substanzen, die die neuronale Morphologie verändern und die neuronale Funktion beeinflussen können. In dieser Studie wurden die morphologischen Veränderungen von intrazellulär mit Biocytin gefüllten CA1 Pyramidenzellen qualitativ und quantitativ analysiert. Die hippocampalen Schnittkulturen, in denen sich bekanntermaßen Pyramidenzellen ähnlich entwickeln wie in vivo, wurden dazu herangezogen, die Effekte der C3bot Toxin-Applikation auf die Verzweigung der Dendriten sowie Anzahl und Dichte der dendritischen Spines zu untersuchen. Drei Gruppen von Zellen wurden verglichen: Erstens Neurone, die in serumhaltigem Medium inkubiert worden waren, zweitens Nervenzellen, die in einem Medium ohne Serum inkubiert worden waren und drittens Zellen, die unter Serumentzug dem C3bot Toxin ausgesetzt worden waren. Die Inkubation dauerte 14 Tage, während die Dauer der Toxinexposition zwischen vier und sechs Stunden betrug. Mit Hilfe eines Computers wurden zweidimensionale Nachbildungen der biocytin-markierten CA1 Pyramidenzellen erstellt, und die Gesamtlänge der Dendriten, die Anzahl der dendritischen Verzweigungspunkte und die Gesamtzahl und Dichte der dendritischen Spines gemessen und statistisch ausgewertet. Signifikante Unterschiede wurden zwischen der mit C3 Toxin behandelten Gruppe und der serumhaltig inkubierten Kontrollgruppe beobachtet. Diese signifikanten morphologischen Veränderungen traten selektiv an den Apikaldendriten der toxinbehandelten CA1 Pyramidenzellen auf. Aus der Behandlung resultierte eine Reduktion der Anzahl apikaler Verzweigungspunkte, der Anzahl der apikalen Spines, der Gesamtzahl (basal und apikal addiert) der Spines sowie der Gesamtspinedichte. Im Gegensatz dazu ergaben sich keine signifikanten Unterschiede zwischen der toxinbehandelten Gruppe und der ohne Serum inkubierten Kontrollgruppe, obwohl der Serumentzug im Vergleich zur serumhaltig inkubierten Kontrollgruppe die Entwicklung der Zellen beeinflusste. Auf Grundlage der beobachteten Veränderungen können wir schließen, dass die Behandlung mit C3 bot einen starken Einfluss selektiv auf die Morphologie der Apikaldendriten ausübt. Der Mechanismus, der dieser selektiven Empfindlichkeit der Apikaldendriten gegenüber dem C3 bot Toxin zugrunde liegt, wird Gegenstand weiterer Untersuchungen sein.