Morphogens are secreted signalling molecules that are expressed in restricted groups of cells within the developing tissue. From there, they are secreted and travel throughout the target field and form concentration gradients. These concentration profiles endow receiving cells with positional information. A number of experiments in Drosophila demonstrated that the morphogen Decapentaplegic (Dpp) forms activity gradients by inducing the expression of several target genes above distinct concentration thresholds at different distances from the source. This way, Dpp contributes to developmental fates in the target field such as the Drosophila wing disc. Although the tissue distribution as well as the actual shape and size of the Dpp morphogen concentration gradient has been visualized, the cell biological mechanisms through which the morphogen forms and maintains a gradient are still a subject of debate. Two hypotheses as to the dominant mechanism of movement have been proposed that can account for Dpp spreading throughout the Drosophila wing imaginal target tissue: extracellular diffusion and planar transcytosis, i. e. endocytosis and resecretion of the ligand that is thereby transported through the cells. Here, I present data indicating that implications of a theoreticalanalysis of Dpp spreading, where Dpp transport through the target tissue is solely based on extracellular diffusion taking into account receptor binding and subsequent internalization, are inconsistent with experimental results. By performing Fluorescence Recovery After Photobleaching (FRAP) experiments, I demonstrate a key role of Dynamin-mediated endocytosis for Dpp gradient formation. In addition, I show that most of GFP-Dpp traffics through endocytic compartments at the receiving epithelial cells, probably recycled through apical recycling endosomes (ARE). Finally, a Dpp recycling assay based on subcellular photouncage of ligand is presented to address specifically the Dpp recycling event at the receiving cells.

Cette thèse est consacrée à certains modèles mathématiques décrivant le mouvement d'une population de cellules, qui interagissent via un signal chimique. L'accent est mis sur le modèle parabolique de Patlak-Keller-Segel, et dans une moindre mesure, sur le modèle cinétique d'Othmer-Dunbar-Alt.<br /><br />Dans une première partie nous étudions plusieurs variantes du modèle PKS classique, incluant notamment une diffusion non-linéaire des cellules, ou bien une loi de diffusion chimique à noyau de Green logarithmique. Puis nous montrons l'existence globale pour une masse sous-critique du modèle PKS classique dans tout l'espace $\mathbb{R}^2$.<br />On complexifie ensuite le modèle de base en ajoutant un intermédiaire chimique réactionnel, ce qui modifie l'homogénéité du système. Enfin les conditions d'existence globale pour le modèle cinétique ODA avec effets délocalisants sont affaiblies par rapport aux travaux précédents.<br /><br />Dans une deuxième partie nous appliquons le modèle phénoménologique de PKS, et son principe de masse critique, à un processus d'auto-organisation remarquable dans le cerveau: la sclérose concentrique de Baló. Un couplage adéquat entre un front de propagation et une instabilité de PKS décrit raisonnablement les motifs en anneaux de la maladie.<br /><br />La troisième partie adopte le point de vue du transport optimal de masse pour analyser le modèle de PKS unidimensionnel modifié auparavant (afin de partager les caractéristiques de PKS 2D). Bien que la fonctionnelle d'énergie ne soit pas convexe par déplacement, nous démontrons la convergence vers un unique état d'équilibre, lorsqu'il existe. Ces nouvelles idées sont mises en oeuvre numériquement~: un flot gradient discret pour la distance de Wasserstein est analysé, puis simulé en dimension un d'espace.<br /><br />Plusieurs annexes viennent compléter ce travail, dont une annexe qui regroupe tous les aspects numériques de la thèse.

Many cutting-edge computer and electronic products are powered by advanced Systems-on-Chip (SoC). Advanced SoCs encompass superb performance together with large number of functions. This is achieved by efficient integration of huge number of transistors. Such very large scale integration is enabled by a core-based design paradigm as well as deep-submicron and 3D-stacked-IC technologies. These technologies are susceptible to reliability and testing complications caused by thermal issues. Three crucial thermal issues related to temperature variations, temperature gradients, and temperature cycling are addressed in this thesis. Existing test scheduling techniques rely on temperature simulations to generate schedules that meet thermal constraints such as overheating prevention. The difference between the simulated temperatures and the actual temperatures is called temperature error. This error, for past technologies, is negligible. However, advanced SoCs experience large errors due to large process variations. Such large errors have costly consequences, such as overheating, and must be taken care of. This thesis presents an adaptive approach to generate test schedules that handle such temperature errors. Advanced SoCs manufactured as 3D stacked ICs experience large temperature gradients. Temperature gradients accelerate certain early-life defect mechanisms. These mechanisms can be artificially accelerated using gradient-based, burn-in like, operations so that the defects are detected before shipping. Moreover, temperature gradients exacerbate some delay-related defects. In order to detect such defects, testing must be performed when appropriate temperature-gradients are enforced. A schedule-based technique that enforces the temperature-gradients for burn-in like operations is proposed in this thesis. This technique is further developed to support testing for delay-related defects while appropriate gradients are enforced. The last thermal issue addressed by this thesis is related to temperature cycling. Temperature cycling test procedures are usually applied to safety-critical applications to detect cycling-related early-life failures. Such failures affect advanced SoCs, particularly through-silicon-via structures in 3D-stacked-ICs. An efficient schedule-based cycling-test technique that combines cycling acceleration with testing is proposed in this thesis. The proposed technique fits into existing 3D testing procedures and does not require temperature chambers. Therefore, the overall cycling acceleration and testing cost can be drastically reduced. All the proposed techniques have been implemented and evaluated with extensive experiments based on ITC’02 benchmarks as well as a number of 3D stacked ICs. Experiments show that the proposed techniques work effectively and reduce the costs, in particular the costs related to addressing thermal issues and early-life failures. We have also developed a fast temperature simulation technique based on a closed-form solution for the temperature equations. Experiments demonstrate that the proposed simulation technique reduces the schedule generation time by more than half.

Cette thèse a pour cadre le développement de méthodes numériques pour la conception optimale de forme de voilure en aérodynamique compressible. Ce travail a donné lieu au développement et à la validation d'un adjoint discret aéro-structure pour l'analyse de sensibilité par rapport aux paramètres de forme et de structure interne de l 'aile dont dépend la fonction d'intérêt, qu' elle soit aérodynamique ou structurale. Les développements logiciels ont été réalisés dans le code de simulation numérique de mécanique des fluides elsA et font suite aux travaux de Marcelet portant sur l'adjoint aéroélastique et dont ils sont une extension. Alors que pour l'adjoint aéro-élastique, on considère une aile flexible, de caractéristiques structurales constantes, pour l'adjoint aérostructure, leur variations sont prises en compte. Pour cela, l'extension de la méthode adjointe s' est accompagnée du développement d' un module de modélisation de la structure interne de l'aile. Ce module est linéarisé et vient donc alimenter le système adjoint. Il a été validé par le dimensionnement de la structure primaire d'une configuration de recherche fournie par l' avionneur Airbus. Dans l'état actuel de développement de la méthode adjointe dans le code elsA, on peut donc désormais calculer les sensibilités d'une fonction aérodynamique ou d'une fonction structure par rapport aux paramètres de forme aérodynamique ou de structure interne de l'aile. Le calcul des gradients ainsi obtenus a été validé par des comparaisons systématiques aux différences finies.

L'objectif de cette thèse est l'accélération des solveurs de Gradient Conjugué avec déflation utilisés pour la résolution de l'équation de Poisson pour la pression, dans le cas de la simulation d'écoulements à faible nombre de Mach sur des maillages non structurés. Une méthode de redémarrage basée sur une estimation de l'effet des erreurs numériques a été mise en œuvre et validée. Par la suite, une méthode à trois niveaux de maillage a été créée, et deux techniques ont dû être développées pour réduire le nombre d'itérations sur les niveaux grossiers : l'une permet la création de solutions initiales grâce à une méthode de projection adaptée, l'autre consiste en une adaptation du critère de convergence sur les niveaux grossiers. Les résultats numériques sur des simulations massivement parallèles montrent entre autres une réduction considérable du temps de calcul global. D'autres pistes de recherche sont introduites, notamment concernant l'équilibrage dynamiques de charge de calcul.

In dimension one, optimal transportation is rather straightforward. The easiness with which a solution can be obtained in that setting has recently been used to tackle more general situations, each time thanks to the same method. First, disintegrate your problem to go back to the unidimensional case, and apply the available 1D methods to get a first result; then, improve it gradually using some evolution process.This dissertation explores that direction more thoroughly. Looking back at two problems only partially solved this way, I show how this viewpoint in fact allows to go even further.The first of these two problems concerns the computation of Yann Brenier's optimal map. Guillaume Carlier, Alfred Galichon, and Filippo Santambrogio found a new way to obtain it, thanks to an differential equation for which an initial condition is given by the Knothe--Rosenblatt rearrangement. (The latter is precisely defined by a series of unidimensional transformations.) However, they only dealt with discrete target measures; I~generalize their approach to a continuous setting. By differentiation, the Monge--Ampère equation readily gives a PDE satisfied by the Kantorovich potential; but to get a proper initial condition, it is necessary to use the Nash--Moser version of the implicit function theorem.The basics of optimal transport are recalled in the first chapter, and the Nash--Moser theory is exposed in chapter 2. My results are presented in chapter 3, and numerical experiments in chapter 4.The last chapter deals with the IDT algorithm, devised by François Pitié, Anil C. Kokaram, and Rozenn Dahyot. It builds a transport map that seems close enough to the optimal map for most applications. A complete mathematical understanding of the procedure is, however, still lacking. An interpretation as a gradient flow in the space of probability measures is proposed, with the sliced Wasserstein distance as the functional. I also prove the equivalence between the sliced and usual Wasserstein distances.

On s'intéresse à la géométrie globale et asymptotique de certaines variétés riemanniennes non compactes. Dans une première partie, on étudie la topologie et la géométrie à l'infini des variétés riemanniennes à courbure (de Ricci) positive ayant un rapport asymptotique de courbure fini. On caractérise le cas non effondré via la notion de cône asymptotique et on donne des conditions suffisantes sur le groupe fondamental pour garantir un non effondrement. La seconde partie est dédiée à l'étude des solutions de Type III du flot de Ricci à courbure positive et aux solitons gradients de Ricci expansifs (points fixes de Type III) présentant une décroissance quadratique de la courbure. On montre l'existence et l'unicité des cônes asymptotiques de tels points fixes. On donne également des conditions suffisantes de nature algébrique et géométrique pour garantir une géométrie de révolution de tels solitons. Dans une troisième partie, on caractérise la géométrie des solitons gradients de Ricci stables à courbure positive et à croissance volumique linéaire. Puis, on s'intéresse au non effondrement des variétés riemanniennes de dimension trois à courbure de Ricci positive ayant un rapport asymptotique de courbure fini. / We study the global and asymptotic geometry of non-compact Riemannian manifolds. First, we study the topology and geometry at infinity of Riemannian manifolds with nonnegative (Ricci) curvature and finite asymptotic curvature ratio. We focus on the non-collapsed case with the help of asymptotic cones and we give sufficient conditions on the fundamental group to guarantee non-collapsing. The second part is dedicated to the study of (non-negatively curved) Type III Ricci flow solutions. We mainly analyze the asymptotic geometry of Type III self-similar solutions (expanding gradient Ricci soliton) with finite asymptotic curvature ratio. We prove the existence and uniqueness of their asymptotic cones. We also give algebraic and geometric sufficient conditions to guarantee rotational symmetry of such metrics. In the last part, we characterize the geometry of steady gradient Ricci solitons with nonnegative sectional curvature and linear volume growth. Finally, we study the non-collapsing of three dimensional Riemannian manifold with nonnegative Ricci curvature and finite asymptotic curvature ratio.

Esta tese apresenta soluções para o gerenciamento e otimização de temperatura para sistemas de codificação de vídeo baseados nas características da aplicação e no conteúdo dos vídeos digitais. Diferente dos trabalhos estado-da-arte, as soluções propostas nesta tese focam em técnicas de gerenciamento de temperatura no nível da aplicação e características da aplicação codificação de vídeo e as propriedades dos vídeos digitais são explorados para desenvolver soluções termais para a codificação de vídeo com baixas perdas na qualidade de serviço das aplicações. Diversas análises são realizadas considerando a aplicação de codificação de vídeo para entender o comportamento da temperatura durante o processo de codificação para diferentes sequências de vídeo. Com base nos resultados das análises, soluções com diferentes abordagens são propostas para atenuar os efeitos da temperatura nos sistemas de codificação de vídeo. Gerenciamento de temperatura baseado nas características da aplicação para o padrão de codificação HEVC usa uma técnica de seleção de configuração em tempo de execução para manter a temperatura abaixo dos limites seguros de operação com bons resultados de qualidade de vídeo. Otimização de temperatura baseado em computação imprecisa usa aproximações baseadas em conteúdo para reduzir a temperatura de chips executando o HEVC. Um escalonador de tarefas que usa características da aplicação para guiar o escalonamento de threads focando na redução dos gradientes espaciais de temperatura que são resultantes do desbalanceamento natural de cargas entre as threads da aplicação. As soluções propostas são capazes de reduzir em até 10 ºC a temperatura do chip com perdas insignificantes na eficiência de compressão. Os resultados de qualidade objetiva (medida usando PSNR) são de 12 dBs até 20 dBs maiores quando comparados com trabalhos da literatura. Além disso, o escalonador de tarefas proposto é capaz de eliminar os gradientes espaciais de temperatura maiores que 5 ºC para arquitetura multi-cores. Como principal conclusão, esta tese demonstra que as técnicas de gerenciamento de temperatura que usam o conhecimento da aplicação de maneira conjunta com as propriedades dos vídeos digitais tem um alto potencial para melhorar os resultados de temperatura de sistemas de codificação de vídeo mantendo bons resultados de qualidade visual dos vídeos codificados. / This thesis presents application-driven temperature-aware solutions for next generation video coding systems, such as the High Efficiency Video Coding (HEVC). Different from state-of-the-art works, the proposed solutions raise the abstraction of temperature management to the application-level, where video coding characteristics and video content properties are used to leverage thermal-aware solutions for video coding with low QoS (Quality of Service) degradation. Several video coding and temperature analyses are performed to understand the behavior of temperature when encoding different video sequences. Based on the analyses results, different approaches are proposed to mitigate the temperature effects on video coding systems. Application-driven temperature management for HEVC uses run-time encoder configuration selection to keep temperature under safe operational state while providing good visual quality results. Temperature optimization using approximate computing uses content-driven approximations to reduce the on-chip temperature of HEVC encoding. Application-driven temperature-aware scheduler leverages application-specific knowledge to guide a scheduling technique targeting reducing the spatial temperature gradients that are resulted from the unbalance workload nature of multi-threaded video coding application. The proposed solutions are able to provide up to 10 °C of chip temperature reduction with negligible compression efficiency loss. Besides, when compared with previous works the resulted objective video quality (PSNR) is from 12 dB up to 20 dB higher. Moreover, the proposed scheduler eliminates spatial temperature gradients greater than 5 ºC of multi-core architectures. As conclusion, this thesis demonstrates that leveraging application-specific knowledge and video content properties has a significant potential to improve temperature profiles of video coding systems while still keeping good quality results.

Magister Scientiae - MSc (Computer Science) / A machine translation system that can convert South African Sign Language (SASL) video to audio or text and vice versa would be bene cial to people who use SASL to communicate. Five fundamental parameters are associated with sign language gestures, these are: hand location; hand orientation; hand shape; hand movement and facial expressions. The aim of this research is to recognise facial expressions and to compare both feature descriptors and machine learning techniques. This research used the Design Science Research (DSR) methodology. A DSR artefact was built which consisted of two phases. The rst phase compared local binary patterns (LBP), compound local binary patterns (CLBP) and histogram of oriented gradients (HOG) using support vector machines (SVM). The second phase compared the SVM to arti cial neural networks (ANN) and random forests (RF) using the most promising feature descriptor|HOG|from the rst phase. The performance was evaluated in terms of accuracy, robustness to classes, robustness to subjects and ability to generalise on both the Binghamton University 3D facial expression (BU-3DFE) and Cohn Kanade (CK) datasets. The evaluation rst phase showed HOG to be the best feature descriptor followed by CLBP and LBP. The second showed ANN to be the best choice of machine learning technique closely followed by the SVM and RF.

The influence of small gradients of altitude (0-500 meters) and order (1st to 4th) in the spatial distribution of mollusk communities was studied, based on 40 samples collected in a watershed of a temperate climate region, in Brazil. Richness was higher in lower ranges of altitude (0-100), and increased with the increase of orders, independently of altitude. However, composition and proportional abundance of species of the communities were not structured by the environmental gradients. The small influence of landscape factors on the spatiality of many abiotic factors, and the tolerance of many species to variations in the environmental factors analyzed determined these results. The proximity of the lower of altitudes with a transitional relief area, and the large width of the 4th order streams, which favored a higher number of ecological niches, provided higher complexity to the sites localized in these areas, allowing the occurrence of higher richness. The existence of patchy spatial structured pattern in good environmental conditions determined the absence of the proportional abundance of species. However, few landscape heterogeneity and possibly, interference of vertebrate vectors in the dispersion of the freshwater mollusks hinder to find clear patterns of spatial distribution. / A influência de pequenos gradientes de altitude (0-500 m) e ordem (1ª a 4ª) na distribuição espacial das comunidades de moluscos foi estudada, com base em 40 amostras coletadas em uma bacia hidrográfica, localizada em região de clima temperado, no Brasil. A riqueza foi maior nas faixas mais baixas de altitude (0-100), e aumentou com o aumento das ordens, independentemente da altitude. No entanto, a composição e abundância proporcional de espécies das comunidades não foram estruturadas pelos gradientes ambientais. A pequena influência dos fatores de paisagem analisados sobre a espacialidade de muitos fatores abióticos, e a tolerância de muitas espécies às variações dos fatores ambientais analisados, determinaram esses resultados. A localização das áreas de altitude mais baixas em uma região de transição de relevo e a maior largura dos rios de 4ª ordem, a qual favorece a existência de um maior número de nichos ecológicos, determinaram maior heterogenidade ambiental, permitindo a ocorrência de maior riqueza. A existência de locais com condições ambientais favoráveis a cada espécie em particular sem estruturação espacial refletiu-se na falta de estruturação das comunidades, quanto a abundância proporcional de espécies. No entanto, a poucas heterogeneidade espacial e, possivelmente, a interferência de vetores vertebrados na dispersão dos moluscos de água doce, impediram encontrar padrões mais claros de distribuição espacial.