Ces travaux ont visé à étudier l’impact de la disponibilité glucidique sur la performance et les adaptations de l’entraînement. L’étude n°1 a permis de caractériser la stratégie de récupération optimale entre deux entraînements chez des athlètes élites en BMX. Les résultats ont permis d’identifier la stratégie nutritionnelle comme étant la stratégie de récupération induisant la plus faible diminution de puissance lors de tests de performance. La deuxième partie a investigué l’impact de la périodisation chronique de l’apport glucidique, par la stratégie « Sleep-Low », sur la performance en endurance et la fonction immunitaire. La stratégie « Sleep-Low» consiste à réaliser des entraînements ciblés en condition de faible disponibilité en glucides, notamment une période récupération faible en glucides et un entraînement à faible intensité réalisé à jeun. Les résultats révèlent que tous les sujets de la stratégie « Sleep-Low » améliorent leur performance en endurance lors d’un 10km en course à pied ou d’un contre-la-montre de 20km. Cette amélioration de la performance est associée à une modification de la stratégie d’allure dans le cas d’un contre-la-montre vers une augmentation des niveaux de puissance développés. Les résultats ont suggéré une modification de la perception d’effort suite à l’implémentation de la stratégie. Ces résultats sont associés à une diminution de la masse grasse chez les participants. L’étude n°3 confirme que la stratégie « Sleep-Low » n’entraîne pas de dégradation de la fonction immunitaire ou de l’efficacité de sommeil. Ces travaux vont dans le sens d’une manipulation de la disponibilité en glucides en fonction de l’objectif des périodes d’entraînement / The aim of the present work is to study the impact of the availability of carbohydrates on performance and on the development of training adaptations. The first section of this thesis aimed at defining the optimal recovery strategies between training sessions for elite BMX pilots. The results revealed that the nutritional strategy is one of the most efficient recovery strategies leading to a lesser decrease in power output. The second section of this thesis investigated the impact of a chronic periodization of carbohydrate intake, through the “Sleep-Low” strategy, on endurance performance and immune function. The “Sleep-Low” strategy consists in specific training sessions performed under low carbohydrate availability, notably during recovery period and low intensity prolonged sessions. All subjects training under the “Sleep-Low” strategy improved their performance on a 10km running trial and on a 20km cycling time trial. This improved performance is associated with a modification of pacing strategy during the 20km cycling time trial toward higher power output. The results suggested a modification of perception of effort after training under the “Sleep-Low” strategy: despite an improvement of performance, participants did not perceive the effort as more difficult. These results are associated to a decrease in fat mass. The study #3 confirmed that the periodization of carbohydrate intake did not impair immune function or sleep efficacy. All the results of the four studies bring new elements in the evolution of carbohydrate recommendations. These works are in the line of a manipulation of carbohydrate availability according to the work required

L'information géographique peut être perçue selon trois dimensions : une dimension spatiale (e.g., ville), une dimension temporelle (e.g., date) et une dimension alphanumérique (e.g., population). L'intégration de ces différents types de données est l'un des challenges actuels de la production d'applications permettant la représentation et le traitement d'informations géographiques. Nous nous proposons de contribuer à cette problématique en proposant une méthode afin d'aider à la gestion de la cohérence des données impliquant deux des trois dimensions. Cette cohérence peut se poser après application d'un opérateur sur les informations géographiques. La méthode que nous proposons est basée sur des liens sémantiques entre les dimensions. Une liaison sémantique représente le rapport logique entre les dimensions. Cette liaison associée à des règles de comportement des opérateurs de manipulation, nous permet de garantir la cohérence du modèle de données associé au résultat d’un opérateur. / Geographic information can be perceived according to three dimensions: a spatial dimension (e.g., city), a temporal dimension (e.g., date) and an alphanumeric dimension (e.g., population). The integration of these various types of data is one of the current challenges in order to model geographic representation and to process geographic information. We contribute to this challenge by proposing a method to help in the management of data coherency involving two on three dimensions. This coherency can be settled after applying an operator on geographic information. Our method is based on semantic links between dimensions. A semantic link represents the logical interaction between (two) dimensions. This connection, associated with a set of rules depending on data manipulation operators, allows us to guarantee the coherency of the data model associated with the result of an operator.

La réalité virtuelle est une technologie qui voit ses applications s’étendre à de nombreux domaines (médical, automobile, etc.). Cette thèse se place dans le contexte des espaces virtuels complètement immersifs, et a pour but d’étudier les effets des deux principaux types d’interfaces proposés (manette avec 6 degrés de liberté, et système de suivi de doigts) sur l’expérience des utilisateurs, dans le cadre de la manipulation d’objets 3D. Nous nous intéressons à des paramètres tels que la facilité d’utilisation, la sensation d’immersion, la rapidité et la précision offertes... Pour cela, nous proposons des expériences évaluant ces paramètres à travers des tâches dont le succès est mesurable, et qui ne sont pas spécifiques à un domaine. Dans une première étude, nous nous intéressons aux tâches complexes d’ordre général, faisant appel à des compétences requises dans les manipulations quotidiennes, telles que le fait d’attraper, de relâcher, de translater, de tourner et de maintenir en équilibre des objets tout en se déplaçant. Nous affinons ensuite notre étude en observant les effets de ces interfaces sur les mouvements eux-mêmes, en les décomposant en degrés de liberté individuels et groupés. Enfin, nous testons l’applicabilité de notre système de manipulation directe dans le cadre d’une étude préliminaire sur l’utilisation de la réalité virtuelle pour le traitement de la maladie d’Alzheimer. Ces études analysent les propriétés de ces interfaces dans le but de fournir des indications aidant au choix de l’interface la plus appropriée pour des applications futures. / Virtual reality is a technology with applications in numerous fields (medical, automotive, etc.). This thesis focuses on immersive virtual spaces, and aims at studying the effects of the two major types of interfaces proposed (6 degree of freedom flystick, and finger-tracking system) on the user experience, and specifically for 3D object manipulation. We are interested in parameters such as ease of use, sense of presence, speed and precision offered. To do so, we design experiments to evaluate these parameters via tasks with measurable success, and which are not field specific.In a first experiment, we study complex general purpose tasks, combining skills required in everyday manipulations, such as grabbing, releasing, translating, rotating, and balancing objects while walking. We then refine our study by observing the effects of those interfaces on the movements themselves, by decomposing them into individual and grouped degrees of freedom. Lastly, we evaluate the applicability of our direct manipulation system in the context of a preliminary study on the use of virtual reality for the treatment of Alzheimer’s disease. These studies analyze the properties of these interfaces to provide guidelines to the choice of the most appropriate interface for future experiments.

Kyrö, F. & Loodin, D. Bedrägeribrott mot Sveriges äldre invånare. Tillvägagångssätt & säkerhetsåtgärder. Examensarbete i Kriminologi, 15 högskolepoäng. Malmö universitet: Fakulteten för hälsa och samhälle, institutionen för Kriminologi, 2022.   Bedrägeribrott mot äldre fortsätter att utgöra ett högst relevant problem i samband med den fortsatta digitaliseringen av vårt samhälle. I takt med att allt fler tjänster blir digitala är det viktigt att vi värnar för att dessa personer inkluderas i utvecklingen. Genom semistrukturerade intervjuer undersöktes vilka tillvägagångssätt som används vid bedrägeribrott mot äldre samt vilka säkerhetsåtgärder som används i dagsläget för att minska deras utsatthet för denna typ av brott. Ur resultatet framkom att befogenhetsbedrägeri tillhör den vanligaste typen av bedrägeribrott mot äldre och att flera olika metoder används. Tekniker som stress och social manipulation samt utnyttjande av social isolering, tillit till samhället och myndigheter samt teknikovana är återkommande i majoriteten av de metoder som används. Vidare framkom det från respondenterna att insatser och åtgärder vars syfte är informationsspridning och beteendeförändring ansågs utgöra de viktigaste proaktiva säkerhetsåtgärderna för att minska risken för fortsatt viktimisering av denna grupp. De främsta reaktiva säkerhetsåtgärderna som lyftes av respondenterna var bankernas säkerhetssystem som larmar om avvikande transaktioner upptäcks samt att anmäla bedrägeriförsök och fullbordade bedrägeribrott till Polisen. Under arbetets gång identifierades även ett omfattande mörkertal gällande antalet äldre som blivit utsatta för bedrägeribrott. Två eventuellt bidragande faktorer till det stora mörkertalet var enligt respondenterna låg anmälningsbenägenheten bland äldre till följd av en skam kopplat till deras utsatthet samt att majoriteten av misslyckade bedrägeriförsöken inte anmäls till Polisen. / Kyrö, F. & Loodin, D. Fraud Against Elderly Citizens in Sweden. Methods, Techniques & Safety measures. Bachelor's thesis in Criminology 15 credits. Malmö University: Faculty of Health and Society, Department of Criminology, 2022.   Fraud against elders remains a pertinent problem as the digitalization of our society continue. As more services become digital it is of outmost importance that we ensure that these people are being included in the progress. By conducting semi structured interviews information were gathered regarding methods and techniques commonly used by culprits of fraud against elderly citizens in Sweden as well as the kind of security measures that are currently in place to prevent further victimization among the population. The results shows that the most common type of fraud perpetrated against elderly citizens in Sweden is courier fraud where the culprit pretends to represent for example a police officer or bank officials. Several different methods are being used but techniques such as the use of stress, social manipulation and exploitation of the victim’s social isolation, trust in society and authorities, and lesser digital skill is reoccurring in most of the methods being used. Measures that aimed to contribute to the spread of information and behavioural change were thought of as the most important proactive security measures. Among the most important reactive security measures were the banks security systems that alert the banks when it detects any abnormal transactions and reporting the crime to the Police authority. During the study a substantial issue regarding the high number of unreported cases. Two potential factors contributing to this was according to the respondent’s low inclination to report due to feelings of shame and that failed attempts of fraud rarely are being reported to the authorities.

In this article, based on Dissection-Motion-Operations, DMO (decomposing a figure into several pieces and composing the resulting pieces into a new figure of equal area), a set of visual representations (models) of mathematical concepts will be introduced. The visual models are producible through manipulation and computer GSP/Cabri software. They are based on the van Hiele’s Levels (van Hiele, 1989) of Thought Development; in particular, Level 2 (Informal Deductive Reasoning) and level 3 (Deductive Reasoning). The basic theme for these models has been visual learning and understanding through manipulatives and computer representations of mathematical concepts vs. rote learning and memorization. The three geometric transformations or motions: Translation, Rotation, Reflection and their possible combinations were used; they are illustrated in several texts. As well, a set of three commonly used dissections or decompositions (Eves, 1972) of objects was utilized.

Comme les robots effectuent de plus en plus de tâches en interaction avec l'homme ou dans un environnement humain, ils doivent assurer la sécurité et le confort des hommes. Dans ce contexte, le robot doit adapter son comportement et agir en fonction des évolutions de l'environnement et des activités humaines. Les robots développés sur la base de l'apprentissage ou d'un planificateur de mouvement ne sont pas en mesure de réagir assez rapidement, c'est pourquoi nous proposons d'introduire un contrôleur de trajectoire intermédiaire dans l'architecture logicielle entre le contrôleur bas niveau et le planificateur de plus haut niveau. Le contrôleur de trajectoire que nous proposons est basé sur le concept de générateur de trajectoire en ligne (OTG), il permet de calculer des trajectoires en temps réel et facilite la communication entre les différents éléments, en particulier le planificateur de chemin, le générateur de trajectoire, le détecteur de collision et le contrôleur. Pour éviter de replanifier toute une trajectoire en réaction à un changement induit par un humain, notre contrôleur autorise la déformation locale de la trajectoire et la modification de la loi d'évolution pour accélérer ou décélérer le mouvement. Le contrôleur de trajectoire peut également commuter de la trajectoire initiale vers une nouvelle trajectoire. Les fonctions polynomiales cubiques que nous utilisons pour décrire les trajectoires fournissent des mouvements souples et de la flexibilité sans nécessiter de calculs complexes. De plus, les algorithmes de lissage que nous proposons permettent de produire des mouvements esthétiques ressemblants à ceux des humains. Ce travail, mené dans le cadre du projet ANR ICARO, a été intégré et validé avec les robots KUKA LWR de la plate-forme robotique du LAAS-CNRS. / In order to perform a large variety of tasks in interaction with human or in human environments, a robot needs to guarantee safety and comfort for humans. In this context, the robot shall adapt its behavior and react to the environment changes and human activities. The robots based on learning or motion planning are not able to adapt fast enough, so we propose to use a trajectory controller as an intermediate control layer in the software structure. This intermediate layer exchanges information with the low level controller and the high level planner. The proposed trajectory controller, based on the concept of Online Trajectory Generation (OTG), allows real time computation of trajectories and easy communication with the different components, including path planner, trajectory generator, collision checker and controller. To avoid the replan of an entire trajectory when reacting to a human behaviour change, the controller must allow deforming locally a trajectory or accelerate/decelerate by modifying the time function. The trajectory controller must also accept to switch from an initial trajectory to a new trajectory to follow. Cubic polynomial functions are used to describe trajectories, they provide smoothness, flexibility and computational simplicity. Moreover, to satisfy the objective of aesthetics, smoothing algorithm are proposed to produce human-like motions. This work, conducted as part of the ANR project ICARO, has been integrated and validated on the KUKA LWR robot platform of LAAS-CNRS.

No description available.

This study seeks to analyze and understand the Russian propaganda before and after the full-scale invasion of Ukraine. 15th of October 2021 to the 28th of February 2022 is the chosen research period, mainly because of the importance to get an understanding of how Kremlin used its propaganda machine to first justify the invasion and then legitimizing it after 24th of February. The study aims to shed light on how the propaganda can be understood from the perspective of the theories classical realism and authoritarian propaganda. Also, from a more general perspective to recognize what truly pervades the Russian propaganda during the period. The result reveals that the propaganda is strongly shaped by realistic beliefs regarding Russia’s national security in relation to the supposed threat coming from Kyiv and the West. It also shows that the propaganda is heavily restricted to the Russian authorities, seeing that it aims to create a reality where the western hegemony is in the wrong, while the Russian society is an idealistic alternative. Therefor the propaganda creates an image that Russia did not have any other choice than to invade Ukraine, to save it from Nazism and the presumed disastrous western civilization.

Le microbiote intestinal joue un rôle important dans la santé des chiens. Les changements dans la composition du microbiote conduisent au déséquilibre de ces micro-organismes qui est appelé dysbiose. Les objectifs de cette étude étaient d’évaluer l’impact de l’administration orale de Saccharomyces cerevisiae sur le microbiote fécal des chiens en bonne santé et d’évaluer le potentiel de Saccharomyces cerevisiae dans la prévention de la dysbiose induite par les antibiotiques. Les chiens ont été divisés en un groupe témoin (n=10) et un groupe probiotique (n=10). Le groupe probiotique a reçu 1 g/kg de S. cerevisiae par jour de D0 à D31. Les deux groupes ont reçu 15 mg/kg de métronidazole par voie orale toutes les 12h, de D11 à D17. Des écouvillons fécaux ont été prélevés sur les échantillons D0, 3, 11, 17, 20, 24 et 31 pour analyse du microbiote. Du sérum sanguin a été prélevé sur D0 et D24 pour des cytokines IL-2, IL-6, IL-8, IL-10, IL-12, IFN-Υ et TNF-α. Le séquençage de l’ADN pour l’analyse du microbiote a été effectué à l’aide de la plateforme Illumina MiSeq et les données ont été analysées à l’aide du logiciel Mothur. La supplémentation en S. cerevisiae a été associée à des changements dans la composition du microbiote après 3 jours (p-value=0,002). Comme on pouvait s’y attendre, le métronidazole a considérablement modifié la composition du microbiote des deux groupes à partir de D11 jusqu’à D17 (valeur p < 0,001). Même si les deux groupes ont changé de façon marquée de D11 à D17, il y avait une différence significative entre les groupes de D17 (p-value=0,012) et de D20 (p-value=0,036), suggérant que le probiotique utilisé a la capacité de moduler le microbiote des chiens confrontés à la dysbiose. Il n’y avait pas de différence significative entre les groupes pour le D24 (p-value=0,388), mais pour le D31, les chiens du groupe probiotique ressemblaient à leur microbiote de référence, tandis que certains animaux du groupe témoin demeuraient dans l’état dysbiotique (p-value=0,002). Le TNF-α avait considérablement diminué dans le groupe des probiotiques à partir de D0 jusqu’à D24 (p-value=0,002). La quantité de TNF-α observée dans le groupe témoin par rapport au groupe probiotique de D24 était également significativement plus élevée (p-value=0,04). Il n’y avait aucune différence significative entre les autres cytokines mesurées dans le sang. On a également observé qu’un sous-ensemble des échantillons de référence (avant la supplémentation en probiotiques) présentait une abondance plus élevée de pathobiogènes (bactéries potentiellement pathogènes comme Escherichia, Helicobacter et Pseudomonadaceae)., tandis que les autres chiens avaient une plus grande abondance d’organismes bénéfiques (tels que Fusobacteriaceae, Bacteroides, Faecalibacillus, Bacterioidaceae, et Ruminococcaceae). Trois jours après la supplémentation en probiotiques, les chiens transportant plus de pathobiogènes se sont rapprochés d’un profil microbiote plus sain. En conclusion, on a observé que l’utilisation de S. cerevisiae était associée à des changements dans la composition du microbiote chez un groupe de chiens malades. Il a également été observé que la supplémentation avec S. cerevisiae a été en mesure de moduler les changements dans le microbiote intestinal pendant la dysbiose induite par les antibiotiques chez les chiens. Mots clés : Saccharomyces cerevisiae, manipulation du microbiote, microbiote intestinal du chien, antibiotiques. / The gut microbiota plays an important role in the health of dogs. The changes in the microbiota composition lead to the imbalance of these microorganisms which is called dysbiosis. The objectives of this study were to evaluate the impact of oral administration of Saccharomyces cerevisiae on the fecal microbiota of healthy dogs and to evaluate the potential of S. cerevisiae in preventing dysbiosis induced by antibiotics. Dogs were divided in a control (n=10) and a probiotic group (n=10). The probiotic group received 1 g/kg of S. cerevisiae per day from D0 to D31. Both groups were given oral metronidazole 15 mg/kg every 12h from D11 to D17. Fecal swabs were collected on D0, 3, 11, 17, 20, 24, and 31 for microbiota analysis. Blood serum was collected on D0 and D24 for measurements of cytokines IL-2, IL-6, IL-8, IL-10, IL-12, IFN-Υ, and TNF-α. DNA sequencing for microbiota analysis was performed using the Illumina MiSeq platform and data was analyzed using the software Mothur. Supplementation with S. cerevisiae was associated with changes in the microbiota composition after 3 days (p-value=0.002). As expected, metronidazole markedly changed the microbiota composition of both groups from D11 to D17 (p-value<0.001). Even though both groups changed markedly from D11 to D17, there was a significant difference between groups on D17 (p-value=0.012) and on D20 (p-value=0.036), suggesting that the probiotic used has the capacity to modulate the microbiota of dogs facing dysbiosis. There was no significant difference between groups on D24 (p-value=0.388) but on D31, dogs from the probiotic group resembled their baseline microbiota while some animals from the control group remained in the dysbiotic state (p-value=0.002). TNF-α was significantly decreased in the probiotic group from D0 to D24 (p-value=0.002). The amount of observed TNF-α in the control group compared to the probiotic group on D24 was also significantly higher (p-value=0.04). There were no significant differences in the other measured cytokines from the blood. It was also observed that a subset of the dogs at baseline (before probiotic supplementation) carried higher abundances of pathobionts (potentially pathogenic bacteria such as Escherichia, Helicobacter, and Pseudomonadaceae), while the other dogs had higher abundances of beneficial organisms (such as Fusobacteriaceae, Bacteroides, Faecalibacillus, Bacterioidaceae, and Ruminococcaceae). Three days after probiotic supplementation, dogs carrying more pathobionts converted into a healthier microbiota profile. In conclusion, the use of S. cerevisiae was associated with beneficial shifts in the microbiota in a group of heathy dogs and the supplementation was able to modulate the dysbiosis caused by the use of antibiotics. Keywords: Probiotics, microbiota manipulation, intestinal dysbiosis, canine microbiome, antibiotics.

This dissertation focuses on applying artificial intelligence techniques to human-robot interaction, which involves human pose estimation and dual-arm robotic manipulation. The motivating application behind this work is autonomous victim extraction in disaster scenarios using a conceptual design of a Semi-Autonomous Victim Extraction Robot (SAVER). SAVER is equipped with an advanced sensing system and two powerful robotic manipulators as well as a head and neck stabilization system to achieve autonomous safe and effective victim extraction, thereby reducing the potential risk to field medical providers. This dissertation formulates the autonomous victim extraction process using a dual-arm robotic manipulation system for human-robot interaction. According to the general process of Human-Robot Interaction (HRI), which includes perception, control, and decision-making, this research applies machine learning techniques to human pose estimation, robotic manipulator modeling, and dual-arm robotic manipulation, respectively. In the human pose estimation, an efficient parallel ensemble-based neural network is developed to provide real-time human pose estimation on 2D RGB images. A 13-limb, 14-joint skeleton model is used in this perception neural network and each ensemble of the neural network is designed for a specific limb detection. The parallel structure poses two main benefits: (1) parallel ensembles architecture and multiple Graphics Processing Units (GPU) make distributed computation possible, and (2) each individual ensemble can be deployed independently, making the processing more efficient when the detection of only some specific limbs is needed for the tasks. Precise robotic manipulator modeling benefits from the simplicity of the controller design and improves the performance of trajectory following. Traditional system modeling relies on first principles, simplifying assumptions and prior knowledge. Any imperfection in the above could lead to an analytical model that is different from the real system. Machine learning techniques have been applied in this field to pursue faster computation and more accurate estimation. However, a large dataset is always needed for these techniques, while obtaining the data from the real system could be costly in terms of both time and maintenance. In this research, a series of different Generative Adversarial Networks (GANs) are proposed to efficiently identify inverse kinematics and inverse dynamics of the robotic manipulators. One four-Degree-of-Freedom (DOF) robotic manipulator and one six-DOF robotic manipulator are used with different sizes of the dataset to evaluate the performance of the proposed GANs. The general methods can also be adapted to other systems, whose dataset is limited using general machine learning techniques. In dual-arm robotic manipulation, basic behaviors such as reaching, pushing objects, and picking objects up are learned using Reinforcement Learning. A Teacher-Student advising framework is proposed to learn a single neural network to control dual-arm robotic manipulators with previous knowledge of controlling a single robotic manipulator. Simulation and experimental results present the efficiency of the proposed framework compared to the learning process from scratch. Another concern in robotic manipulation is safety constraints. A variable-reward hierarchical reinforcement learning framework is proposed to solve sparse reward and tasks with constraints. A task of picking up and placing two objects to target positions while keeping them in a fixed distance within a threshold is used to evaluate the performance of the proposed method. Comparisons to other state-of-the-art methods are also presented. Finally, all the three proposed components are integrated as a single system. Experimental evaluation with a full-size manikin was performed to validate the concept of applying artificial intelligence techniques to autonomous victim extraction using a dual-arm robotic manipulation system. / Doctor of Philosophy / Using mobile robots for autonomous victim extraction in disaster scenarios reduces the potential risk to field medical providers. This dissertation focuses on applying artificial intelligence techniques to this human-robot interaction task involving pose estimation and dual-arm manipulation for victim extraction. This work is based on a design of a Semi-Autonomous Victim Extraction Robot (SAVER). SAVER is equipped with an advanced sensing system and two powerful robotic manipulators as well as a head and neck stabilization system attached on an embedded declining stretcher to achieve autonomous safe and effective victim extraction. Therefore, the overall research in this dissertation addresses: human pose estimation, robotic manipulator modeling, and dual-arm robotic manipulation for human pose adjustment. To accurately estimate the human pose for real-time applications, the dissertation proposes a neural network that could take advantages of multiple Graphics Processing Units (GPU). Considering the cost in data collection, the dissertation proposed novel machine learning techniques to obtain the inverse dynamic model and the inverse kinematic model of the robotic manipulators using limited collected data. Applying safety constraints is another requirement when robots interacts with humans. This dissertation proposes reinforcement learning techniques to efficiently train a dual-arm manipulation system not only to perform the basic behaviors, such as reaching, pushing objects and picking up and placing objects, but also to take safety constraints into consideration in performing tasks. Finally, the three components mentioned above are integrated together as a complete system. Experimental validation and results are discussed at the end of this dissertation.