No description available.

No description available.

No description available.

The determination of the solar oxygen abundance is a topic that has long beenunder debate, where the exact value has varied between researchers. These variationsin the oxygen abundance have implications on our understanding of the Sun and itsinterior as well as our understanding of other stars. The aim of this report is to explorethe di erences between three observations; the Liège, Hamburg and SST, and their effecton the solar oxygen abundance calculated from the 777 nm triplet. In the reportit was found that the abundance from the Liège and SST data was lower and in betteragreement (A(O) ≈ 8.709 dex) compared to the Hamburg data (A(O) ≈ 8.732 dex).One of the potential reasons for these di erences explored in the thesis were the e ectsof magnetic  elds. After comparing the di erences it was argued that the SST dataset as well as the Liège could be considered more reliable compared to the Hamburg.The argued abundance was therefore of a somewhat lower value which does not helplessen the inconsistency between solar models and helioseismology. / Bestämningen av mängden syre i solen har länge varit ett debatterat område där det exakta numeriska värdet varierat bland forskare. Dessa variationer av mängden syre ger konsekvenser för vår förståelse av solen och dess inre samt för vår förståelse av andra stjärnor. Syftet med denna rapport är att utforska skillnaderna mellan tre observationer; från Liège, Hamburg och SST, samt deras påverkan på mängden syre i solen beräknad från 777 nm tripletten. Rapporten fann att mängden syre beräknad från Liège- och SST-datan både var lägre samt stämde bättre överens (A(O) ≈ 8.709 dex) i jämförelse med mängden från Hamburg observationen (A(O) ≈ 8.732 dex). En potentiell anledning till skillnaden i resultat som utforskades i denna uppsats var påverkan av magnetiska fält. Efter jämförelse av skillnaderna kunde det konstateras att datan från SST så väl som från Liège kan anses mer pålitliga i förhållande till Hamburg-datan. Den förespråkade mängden syre var åt det lägre hållet vilket inte bidrar till minskningen av avvikelserna mellan modeller av solen och mätningar via helioseismologi.

L’augmentation de la vitesse ferroviaire impose l’amélioration des performances de freinage liés aux matériaux de friction utilisés dans les freins à disques et à semelles. Cependant, l'usure reste un point limitant en terme de performance mais aussi de sécurité. Elle a pour conséquence de détériorer les matériaux (aspects mécano-thermiques) mais aussi de contaminer le contact entre roue et rail par des particules de semelles de frein pouvant isoler électriquement celui-ci (aspects mécano-électriques). Le procédé permettant de localiser les trains, ayant pour principe la conduction électrique entre roue et rail, est alors compromis.Ces problématiques sont gouvernées par les différents contacts (roue/rail, roue/semelle, ...) qui s'inscrivent dans le concept de triplet tribologique. Un triplet est composé des corps en contact (premiers corps), de l'interface (troisième corps) et du mécanisme qui les maintient en contact. Afin de comprendre les phénomènes se produisant au sein des contacts, une analyse multi-physique locale et un découplage des paramètres (mécaniques, thermiques, électriques) s'avèrent nécessaires. Toutefois, cela est difficile à établir expé- rimentalement, une modélisation numérique par éléments discrets est alors pertinente pour palier à ces besoins.L'écoulement dynamique d'un troisième corps constitué de particules de freinage et la dégradation d'un matériau de freinage réel (suite à des méthodes d'imageries) sont modélisés. Les résultats obtenus ont permis d'appréhender les compétitions existant entre paramètres physiques d'une part (mécanique, électrique, thermique) et paramètres d'éch-elles d'autre part (local, global). Ces compétitions tendent à minimiser la résistance élec- trique lorsque les constituants sont en proportion adéquat (meilleure conduction électrique entre roue et rail) mais aussi à équilibrer les débits du circuit tribologique, ce qui conduit à protéger les matériaux (usure contrôlée). En outre, les analyses numériques ont permis de mettre en évidence de nouveaux paramètres liés à ``l'étalement'' du troisième corps et d'appréhender les points clés permettant de se rapprocher d'un contact réel. / The increase of the rail speed imposes the improvement of the braking performances related to the friction materials used in brake pad and brakes shoes. However, wear remains a limiting point in terms of performance but also safety. It has the effect of damaging the materials (mechano-thermal aspects) but also to contaminate the contact between wheel and rail by particles of brake shoes that can electrically isolate it (mechanical-electrical aspects). The method for locating the trains, whose principle is the electrical conduction between wheel and rail, is then compromised.These issues are governed by the different contacts (wheel/rail, wheel/shoes, ...) that fit into the concept of tribological triplets. These are composed of the bodies in contact (first bodies), the interface (third body) and the mechanism that keeps them in contact. In order to understand the phenomena occurring within the contacts, a local multi-physics analysis and a decoupling of the parameters (mechanical, thermal, electrical) are necessary. However, this is difficult to establish experimentally, numerical modeling by discrete elements method is then relevant for these needs.The dynamic flow of a third body consisting of braking particles and the degradation of a real braking material (following imaging methods) are modeled. The results obtained made it possible to understand the competitions existing between physical parameters on the one hand (mechanical, electrical, thermal) and scale parameters on the other hand (local, global). These competitions tend to minimize the electrical resistance when the constituents are in adequate proportion (better electrical conduction between wheel and rail) but also to balance the flows of the tribological track, which leads to protect the materials (controlled wear). In addition, the numerical analyzes made it possible to highlight new parameters related to "spreading" of the third body and to apprehend the key points making it possible to approach a real contact.

Les modèles de Markov à sauts (appelés JMS pour Jump Markov System) sont utilisés dans de nombreux domaines tels que la poursuite de cibles, le traitement des signaux sismiques et la finance, étant donné leur bonne capacité à modéliser des systèmes non-linéaires et non-gaussiens. De nombreux travaux ont étudié les modèles de Markov linéaires pour lesquels bien souvent la restauration de données est réalisée grâce à des méthodes d’échantillonnage statistique de type Markov Chain Monte-Carlo. Dans cette thèse, nous avons cherché des solutions alternatives aux méthodes MCMC et proposons deux originalités principales. La première a consisté à proposer un algorithme de restauration non supervisée d’un JMS particulier appelé « modèle de Markov couple à sauts conditionnellement gaussiens » (noté CGPMSM). Cet algorithme combine une méthode d’estimation des paramètres basée sur le principe Espérance-Maximisation (EM) et une méthode efficace pour lisser les données à partir des paramètres estimés. La deuxième originalité a consisté à étendre un CGPMSM spécifique appelé CGOMSM par l’introduction des copules. Ce modèle, appelé GCOMSM, permet de considérer des distributions plus générales que les distributions gaussiennes tout en conservant des méthodes de restauration optimales et rapides. Nous avons équipé ce modèle d’une méthode d’estimation des paramètres appelée GICE-LS, combinant le principe de la méthode d’estimation conditionnelle itérative généralisée et le principe des moindre-carrés linéaires. Toutes les méthodes sont évaluées sur des données simulées. En particulier, les performances de GCOMSM sont discutées au regard de modèles de Markov non-linéaires et non-gaussiens tels que la volatilité stochastique, très utilisée dans le domaine de la finance. / Switching Markov Models, also called Jump Markov Systems (JMS), are widely used in many fields such as target tracking, seismic signal processing and finance, since they can approach non-Gaussian non-linear systems. A considerable amount of related work studies linear JMS in which data restoration is achieved by Markov Chain Monte-Carlo (MCMC) methods. In this dissertation, we try to find alternative restoration solution for JMS to MCMC methods. The main contribution of our work includes two parts. Firstly, an algorithm of unsupervised restoration for a recent linear JMS known as Conditionally Gaussian Pairwise Markov Switching Model (CGPMSM) is proposed. This algorithm combines a parameter estimation method named Double EM, which is based on the Expectation-Maximization (EM) principle applied twice sequentially, and an efficient approach for smoothing with estimated parameters. Secondly, we extend a specific sub-model of CGPMSM known as Conditionally Gaussian Observed Markov Switching Model (CGOMSM) to a more general one, named Generalized Conditionally Observed Markov Switching Model (GCOMSM) by introducing copulas. Comparing to CGOMSM, the proposed GCOMSM adopts inherently more flexible distributions and non-linear structures, while optimal restoration is feasible. In addition, an identification method called GICE-LS based on the Generalized Iterative Conditional Estimation (GICE) and the Least-Square (LS) principles is proposed for GCOMSM to approximate any non-Gaussian non-linear systems from their sample data set. All proposed methods are tested by simulation. Moreover, the performance of GCOMSM is discussed by application on other generable non-Gaussian non-linear Markov models, for example, on stochastic volatility models which are of great importance in finance.

It has long been recognized that course and efficiency of a chemical reaction is largely mediated by the short-lived transient species (excited state or radicals) which are formed as reactive intermediates during a chemical reaction. Subtle changes not only in the bonding and electronic distributions but also in the conformations and geometries of these intermediates have a dramatic influence on the reactivity. A detailed understanding of the structural and dynamical aspects of electronic excited states is therefore essential towards unraveling photoinduced natural processes and for designing novel photonic materials. Time-resolved techniques have been widely used to study the transient species (or intermediates) formed during photochemical and photophysical reactions for better understanding of the reaction mechanism and dynamics. Time-resolved absorption spectroscopy is a promising tool to study the temporal dynamics and the kinetics of photophysical processes. But the absorption spectra of species in solution usually consist of broad spectral band revealing little or no information about the structure of the transient species under investigation. Time-resolved resonance Raman (TR3) spectroscopy, on the other hand, is a potential sensitive modality, which not only allows one to study the dynamics but also provides the vibrational structure of the transient species of interest in microsecond to picosecond time scale. Moreover, by choosing the wavelength of excitation one can selectively probe the particular transient species from a complex molecular system especially a biological molecule. Thioxanthone (TX) is well known for its dramatic solvatochromic behavior and has drawn enormous attention in the recent years. The objective of present thesis has been to understand the solvent-induced structural changes on the lowest excited triplet state of TX and its derivatives. We have primarily employed nanosecond TR3 spectroscopy, a pump-probe technique, to investigate structure of the lowest excited triplet state. Transient absorption experiments have also been carried out to study the excited electronic states. In order to substantiate our experimental findings and also to get more insight into the triplet-state structure, we have performed density functional theory (DFT) calculations. The polarizable continuum solvation model has been employed to account for the solvent effect into the computation. Time dependent (TD) -DFT calculations have also been performed to get the energy and the structure of the excited states. The present thesis has been divided into eight chapters. Chapter 1 gives brief literature review on photochemistry and photophysics of TX and the introduction to the TR3 technique. In this chapter we have briefly introduced key concepts which form the basis of the thesis. Chapter 2 covers the experimental and theoretical methodologies used in the present thesis work. The major components of the TR3 spectrometer as well as the important technical aspect of the TR3 technique have been discussed in detail. In the section of the theoretical method, basic concepts of the computational method, density functional theory and key concepts related to the solvation are briefly discussed. Chapter 3 focuses on a systematic vibrational study of the ground and lowest triplet states of TX. TR3 experiments have been carried out and the observed vibrational frequencies have been assigned. It has been observed that electronic excitation distorts the molecule, enabling the increased electron delocalization in the central ring keeping the ground state symmetry intact. The largest structural reorganization is observed in the central ring of TX, consisting of an oxygen atom. Normal mode analyses show that the normal mode composition is significantly influenced by the electronic excitation. The C=C stretching and C=O stretching modes are coupled to a greater extent in the triplet state as compared to the ground state. In the ground state, the two high-frequency modes can be assigned almost exclusively to the C=O stretching and C=C stretching, whereas in the triplet state, both of these coordinates have comparable contributions to the two totally symmetric modes. Chapter 4 deals with a very unique observation of simultaneous detection of two triplets. This is the first time when two triplet states have been simultaneously deleted using TR3 experiments. We have performed TR3 experiments in wide variety of solvents differing in their polarities and hydrogen atom donor abilities. The transient Raman signal has been observed from both n - π∗ and π - π∗ triplet states simultaneously. The population ratio of the two triplet states has been found to be dependent on the solvent polarity. Additionally, the excitation wavelength study has revealed that the relative ratios of the transient Raman peaks (assigned to two different triplet states) change with the excitation wavelength. Our claim of simultaneous detection of two triplets has been reconfirmed by triplet quenching experiments carried out at different temperature. It has also been observed that the CO bond length is very sensitive to the solvent polarity and specific interactions play an important role in determining the structure of lowest triplet-state. In Chapter 5, we focus on the understanding of the effect of chlorine substitution on the lowest excited triplet state of TX. TR3 spectroscopy has been used as an experimental tool to study the vibrational structure of 2-chlorothioxanthone (CTX). TR3 results indicate the coexistence of two lowest triplet states in the thermal equilibrium akin to the parent compound. The above observation has been further substantiated by probe wavelength dependent study. The configuration of the T1 state has been assigned to π – π∗, whereas the T2 state has been ascribed as n - π∗. The population ratio of 3n - π∗ to 3 π - π ∗ triplet states has been found to be more for CTX as compared to TX which has been substantiated by the flash photolysis experiments. Chapter 6 highlights the influence of solvent effect on lowest triplet state structure of CTX. Transient absorption spectroscopy has been employed to understand the triplet state electronic structure; whereas solvent induced changes in the structure of the lowest triplet state have been studied using TR3 spectroscopy. Time-resolved absorption measurements show that solvent polarity has dramatic dependence on the wavelength of T1 - Tn absorption maximum. A good correlation between the wavelength of T1 - Tn absorption maximum and ET(30) value of the solvent is observed. TR3 experiments carried out in solvents of varying polarities indicate that the contribution of n - π∗ character to the lowest excited triplet state increases with the increase in the solvent polarity. Both transient absorption and TR3 studies reveal that specific solvent effect is more pronounced in comparison to the nonspecific solvent effect. Chapter 7 of the thesis deals with the study on the triplet state structure and solvent effect on 2-trifluoromethyl Thioxanthone. Flash photolysis in tandem with TR3 spectroscopy has been employed to understand both the electronic and the vibrational structures of this pharmaceutically important thioxanthone derivative. Experiments have also been carried out in solvents of varying polarities to study solvent-induced changes in the triplet-state electronic spectra. We have observed the coexistence of two lowest triplet states alike the parent compound. The T1 state has been assigned to π - π∗ state, whereas n - π∗ configuration has been attributed to the T2 state. The wavelength of triplet-triplet absorption maximum of the lowest triplet state has been found to be sensitive to the solvent polarity and good correlation has been observed with the ET(30) value. The transient Raman results indicate that the CF3 substitution leads to increase in the population ratio of n - π∗ and π - π ∗ triplet states. Finally, Chapter 8 contains overall summary of the thesis and future directions of the present investigation.

Cette thèse est consacrée au problème d'estimation bayésienne pour le filtrage statistique, dont l'objectif est d'estimer récursivement des états inconnus à partir d'un historique d'observations, dans un modèle stochastique donné. Les modèles stochastiques considérés incluent principalement deux grandes classes de modèles : les modèles de Markov cachés et les modèles de Markov à sauts conditionnellement markoviens. Ici, le problème est abordé sous sa forme générale dans la mesure où nous considérons le problème du filtrage mono- et multi objet(s), ce dernier étant abordé sous l'angle de la théorie des ensembles statistiques finis et du filtre « Probability Hypothesis Density ». Tout d'abord, nous nous intéressons à l'importante classe d'approximations que constituent les algorithmes de Monte Carlo séquentiel, qui incluent les algorithmes d'échantillonnage d'importance séquentiel et de filtrage particulaire auxiliaire. Les boucles de propagation mises en jeux dans ces algorithmes sont étudiées et des algorithmes alternatifs sont proposés. Les algorithmes de filtrage particulaire dits « localement optimaux », c'est à dire les algorithmes d'échantillonnage d'importance avec densité d'importance conditionnelle optimale et de filtrage particulaire auxiliaire pleinement adapté sont comparés statistiquement, en fonction des paramètres du modèle donné. Ensuite, les méthodes de réduction de variance basées sur le théorème de Rao-Blackwell sont exploitées dans le contexte du filtrage mono- et multi-objet(s) Ces méthodes, utilisées principalement en filtrage mono-objet lorsque la dimension du vecteur d'état à estimer est grande, sont dans un premier temps étendues pour les approximations Monte Carlo du filtre Probability Hypothesis Density. D'autre part, des méthodes de réduction de variance alternatives sont proposées : bien que toujours basées sur le théorème de Rao-Blackwell, elles ne se focalisent plus sur le caractère spatial du problème mais plutôt sur son caractère temporel. Enfin, nous abordons l'extension des modèles probabilistes classiquement utilisés. Nous rappelons tout d'abord les modèles de Markov couple et triplet dont l'intérêt est illustré à travers plusieurs exemples pratiques. Ensuite, nous traitons le problème de filtrage multi-objets, dans le contexte des ensembles statistiques finis, pour ces modèles. De plus, les propriétés statistiques plus générales des modèles triplet sont exploitées afin d'obtenir de nouvelles approximations de l'estimateur bayésien optimal (au sens de l'erreur quadratique moyenne) dans les modèles à sauts classiquement utilisés; ces approximations peuvent produire des estimateurs de performances comparables à celles des approximations particulaires, mais ont l'avantage d'être moins coûteuses sur le plan calculatoire / This thesis focuses on the Bayesian estimation problem for statistical filtering which consists in estimating hidden states from an historic of observations over time in a given stochastic model. The considered models include the popular Hidden Markov Chain models and the Jump Markov State Space Systems; in addition, the filtering problem is addressed under a general form, that is to say we consider the mono- and multi-object filtering problems. The latter one is addressed in the Random Finite Sets and Probability Hypothesis Density contexts. First, we focus on the class of particle filtering algorithms, which include essentially the sequential importance sampling and auxiliary particle filter algorithms. We explore the recursive loops for computing the filtering probability density function, and alternative particle filtering algorithms are proposed. The ``locally optimal'' filtering algorithms, i.e. the sequential importance sampling with optimal conditional importance distribution and the fully adapted auxiliary particle filtering algorithms, are statistically compared in function of the parameters of a given stochastic model. Next, variance reduction methods based on the Rao-Blackwell theorem are exploited in the mono- and multi-object filtering contexts. More precisely, these methods are mainly used in mono-object filtering when the dimension of the hidden state is large; so we first extend them for Monte Carlo approximations of the Probabilty Hypothesis Density filter. In addition, alternative variance reduction methods are proposed. Although we still use the Rao-Blackwell decomposition, our methods no longer focus on the spatial aspect of the problem but rather on its temporal one. Finally, we discuss on the extension of the classical stochastic models. We first recall pairwise and triplet Markov models and we illustrate their interest through several practical examples. We next address the multi-object filtering problem for such models in the random finite sets context. Moreover, the statistical properties of the more general triplet Markov models are used to build new approximations of the optimal Bayesian estimate (in the sense of the mean square error) in Jump Markov State Space Systems. These new approximations can produce estimates with performances alike those given by particle filters but with lower computational cost

With recent advancements, image generation has become more and more possible with the introduction of stronger generative artificial intelligence (AI) models. The idea and ability of generating non-existing images that highly resemble real world images is interesting for many use cases. Generated images could be used, for example, to augment, extend or replace real data sets for training AI models, therefore being capable of minimising costs on data collection and similar processes. Deep learning, a sub-field within the AI field has been on the forefront of such methodologies due to its nature of being able to capture and learn highly complex and feature-rich data. This work focuses on deep generative learning approaches within a forestry application, with the goal of generating tree log end images in order to enhance an AI model that uses such images. This approach would not only reduce costs of data collection for this model, but also many other information extraction models within the forestry field. This thesis study includes research on the state of the art within deep generative modelling and experiments using a full pipeline from a deep generative modelling stage to a log end recognition model. On top of this, a variant architecture and image sampling algorithm are proposed to add in this pipeline and evaluate its performance. The experiments and findings show that the applied generative model approaches show good feature learning, but lack the high-quality and realistic generation, resulting in more blurry results. The variant approach resulted in slightly better feature learning with a trade-off in generation quality. The proposed sampling algorithm proved to work well on a qualitative basis. The problems found in the generative models propagated further into the training of the recognition model, making the improvement of another AI model based on purely generated data impossible at this point in the research. The results of this research show that more work is needed on improving the application and generation quality to make it resemble real world data more, so that other models can be trained on artificial data. The variant approach does not improve much and its findings contribute to the field by proving its strengths and weaknesses, as with the proposed image sampling algorithm. At last this study provides a good starting point for research within this application, with many different directions and opportunities for future work.

Zero-Shot Learning (ZSL) for image classification aims to recognize images from novel classes for which we have no training examples. A common approach to tackling such a problem is by transferring knowledge from seen to unseen classes using some auxiliary semantic information of class labels in the form of class embeddings. Most of the existing methods represent image features and class embeddings as point vectors, and such vector representation limits the expressivity in terms of modeling the intra-class variability of the image classes. In this thesis, we propose three novel ZSL methods that represent image features and class labels as distributions and learn their corresponding parameters as distribution embeddings. Therefore, the intra-class variability of image classes is better modeled. The first model is a Triplet model, where image features and class embeddings are projected as Gaussian distributions in a common space, and their associations are learned by metric learning. Next, we have a Triplet-VAE model, where two VAEs are trained with triplet based distributional alignment for ZSL. The third model is a simple Probabilistic Classifier for ZSL, which is inspired by energy-based models. When evaluated on the common benchmark ZSL datasets, the proposed methods result in an improvement over the existing state-of-the-art methods for both traditional ZSL and more challenging Generalized-ZSL (GZSL) settings. / Zero-Shot Learning (ZSL) för bildklassificering syftar till att känna igen bilder från nya klasser som vi inte har några utbildningsexempel för. Ett vanligt tillvägagångssätt för att ta itu med ett sådant problem är att överföra kunskap från sett till osynliga klasser med hjälp av någon semantisk information om klassetiketter i form av klassinbäddningar. De flesta av de befintliga metoderna representerar bildfunktioner och klassinbäddningar som punktvektorer, och sådan vektorrepresentation begränsar uttrycksförmågan när det gäller att modellera bildklassernas variation inom klass. I denna avhandling föreslår vi tre nya ZSL-metoder som representerar bildfunktioner och klassetiketter som distributioner och lär sig deras motsvarande parametrar som distributionsinbäddningar. Därför är bildklassernas variation inom klass bättre modellerad. Den första modellen är en Triplet-modell, där bildfunktioner och klassinbäddningar projiceras som Gaussiska fördelningar i ett gemensamt utrymme, och deras föreningar lärs av metrisk inlärning. Därefter har vi en Triplet-VAE-modell, där två VAEs tränas med tripletbaserad fördelningsinriktning för ZSL. Den tredje modellen är en enkel Probabilistic Classifier för ZSL, som är inspirerad av energibaserade modeller. När de utvärderas på de vanliga ZSLdatauppsättningarna, resulterar de föreslagna metoderna i en förbättring jämfört med befintliga toppmoderna metoder för både traditionella ZSL och mer utmanande Generalized-ZSL (GZSL) -inställningar.