No description available.

La recherche d'images basée sur le contenu visuel est un domaine très actif de la vision par ordinateur, car le nombre de bases d'images disponibles ne cesse d'augmenter.L’objectif de ce type d’approche est de retourner les images les plus proches d'une requête donnée en terme de contenu visuel.Notre travail s'inscrit dans un contexte applicatif spécifique qui consiste à indexer des petites bases d'images expertes sur lesquelles nous n'avons aucune connaissance a priori.L’une de nos contributions pour palier ce problème consiste à choisir un ensemble de descripteurs visuels et de les placer en compétition directe. Nous utilisons deux stratégies pour combiner ces caractéristiques : la première, est pyschovisuelle, et la seconde, est statistique.Dans ce contexte, nous proposons une approche adaptative non supervisée, basée sur les sacs de mots et phrases visuels, dont le principe est de sélectionner les caractéristiques pertinentes pour chaque point d'intérêt dans le but de renforcer la représentation de l'image.Les tests effectués montrent l'intérêt d'utiliser ce type de méthodes malgré la domination des méthodes basées réseaux de neurones convolutifs dans la littérature.Nous proposons également une étude, ainsi que les résultats de nos premiers tests concernant le renforcement de la recherche en utilisant des méthodes semi-interactives basées sur l’expertise de l'utilisateur. / Image Retrieval is still a very active field of image processing as the number of available image datasets continuously increases.One of the principal objectives of Content-Based Image Retrieval (CBIR) is to return the most similar images to a given query with respect to their visual content.Our work fits in a very specific application context: indexing small expert image datasets, with no prior knowledge on the images. Because of the image complexity, one of our contributions is the choice of effective descriptors from literature placed in direct competition.Two strategies are used to combine features: a psycho-visual one and a statistical one.In this context, we propose an unsupervised and adaptive framework based on the well-known bags of visual words and phrases models that select relevant visual descriptors for each keypoint to construct a more discriminative image representation.Experiments show the interest of using this this type of methodologies during a time when convolutional neural networks are ubiquitous.We also propose a study about semi interactive retrieval to improve the accuracy of CBIR systems by using the knowledge of the expert users.

Jusqu’à présent, la dynamique des fractions végétales n’avait jamais été étudiée dans les estuaires macrotidaux en raison de leur faible quantité, par rapports aux fortes charges en matières en suspension fines, et du manque de protocoles d’étude et d’échantillonnage adéquats. Les débris végétaux sont toutefois à l’origine de perturbations d’activités économiques qui impliquent la filtration de larges volumes d’eau. L’objectif de cette thèse était ainsi de comprendre la dynamique d’apport et de transit de ces débris végétaux dans l’estuaire de la Gironde par la mise en oeuvre d’un suivi spatio-temporel de leur distribution et de techniques nouvelles pour un tel estuaire hyper-turbide (incubations in-situ litter-bag, caractérisation biogéochimique, identification des sources). Les résultats principaux sont la mise en évidence du contrôle du régime hydrologique sur leur distribution et la détermination des échelles de temps de leur persistance dans l’estuaire de la Gironde. / Up to now, vegetal fraction dynamics has not been studied in macrotidal estuaries, due to its low quantity in comparison to the strong charge of suspended particulate matter, and due to the lack of appropriate protocols to sample and examine it. Nevertheless, vegetal debris have been identified as a factor able to disrupt a wide range of stakeholder activities that require huge volumes of water to filter. The objective of this PhD was to understand the input and transfer dynamics of vegetal debris in the Gironde Estuary by the implementation of a spatiotemporal btrack of their distribution and by the development of new techniques for this hyper-turbide estuary (in situ litter-bag incubations, biogeochemical characterization, and identification of the sources). Mean results highlight the influence of the hydrological regime in their distribution and reveal time scales of their persistence in the Gironde Estuary.

With the availability of inexpensive hardware and software, digital imaging has become an important medium of communication in our daily lives. A huge amount of digital images are being collected and become available through the internet and stored in various fields such as personal image collections, medical imaging, digital arts etc. Therefore, it is important to make sure that images are stored, searched and accessed in an efficient manner. The use of bag of visual words (BOW) model for modelling images based on local invariant features computed at interest point locations has become a standard choice for many computer vision tasks. Based on this promising model, this thesis investigates three main problems: natural scene classification, annotation and retrieval. Given an image, the task is to design a system that can determine to which class that image belongs to (classification), what semantic concepts it contain (annotation) and what images are most similar to (retrieval). This thesis contributes to scene classification by proposing a weighting approach, named keypoints density-based weighting method (KDW), to control the fusion of colour information and bag of visual words on spatial pyramid layout in a unified framework. Different configurations of BOW, integrated visual vocabularies and multiple image descriptors are investigated and analyzed. The proposed approaches are extensively evaluated over three well-known scene classification datasets with 6, 8 and 15 scene categories using 10-fold cross validation. The second contribution in this thesis, the scene annotation task, is to explore whether the integrated visual vocabularies generated for scene classification can be used to model the local semantic information of natural scenes. In this direction, image annotation is considered as a classification problem where images are partitioned into 10x10 fixed grid and each block, represented by BOW and different image descriptors, is classified into one of predefined semantic classes. An image is then represented by counting the percentage of every semantic concept detected in the image. Experimental results on 6 scene categories demonstrate the effectiveness of the proposed approach. Finally, this thesis further explores, with an extensive experimental work, the use of different configurations of the BOW for natural scene retrieval.

Ce travail a porté sur l'étude des instabilités de gradient de température ioniques (ITG) en géométrie cylindrique, le champ magnétique étant supposé constant et dirigé selon l'axe du cylindre. Une fonction de distribution discrète en forme de marche d'escalier est utilisée pour décrire la direction de vitesse parallèle au champ magnétique. L'équation de Vlasov se résume à un système de type multi fluides couplés par l'équation de quasi neutralité. Chaque fluide est décrit par un système fermé d'équations (continuité, Euler et fermeture adiabatique), caractéristiques d'un fluide incompressible, d'où la dénomination de sac d'eau ou "water bag". Le recours à cette description water bag est particulièrement intéressant dans le cas de problèmes à une seule dimension en vitesse. Ainsi, dans le cas des plasmas fortement magnétisés, un modèle water bag peut se combiner avantageusement aux modèles dits girocinétiques. Les paramètres associés a la représentation water bag ont pu être identifiés et reliés aux grandeurs macroscopiques par le biais d'une méthode originale d'équivalence au sens des moments. L'analyse water bag des ITG a permis de valider le modèle et les méthodes choisies. Ce travail a également permis de montrer que le concept de water bag peut sans problème prendre en compte des effets variés comme ceux liés a l'introduction d?un rayon de Larmor fini, tout comme à la description d'un plasma composé de plusieurs espèces d'ions. / A drift-kinetic model in cylindrical geometry has been used to study Ion Temperature Gradients (ITG). The cylindrical plasma is considered as a limit case of a stretched torus. The magnetic field is assumed uniform and constant; it is directed along the axis of the column. A discrete distribution function f taking the form of a multi-step like function is used in place of the continuous distribution function along the parallel velocity direction. With respect to the properties of the Heaviside?s distribution, the Vlasov equation is reduced to a system of fluids coupled by the electromagnetic fields. This model is well suited mainly for problems involving a phase space with one velocity component. In the case of magnetized plasmas it gives an alternative way to study turbulence thanks to the gyro-average whose allows reducing the 3D velocity space into a 1D space. Parameters introduced by the water bag formalism have been linked to physical quantities by an original method of moment-sense equivalence. In the linear approximation, the water bag study of the ITG instability allows an interesting comparison with some well-known analytical results. The water-bag concept is not affected by taking into account Finite Larmor Radius effects. It well describes the case of multi-species plasma

The field of natural language processing has received increased attention lately, but less focus is put on comparing models, which differ in complexity. This thesis compares Random Forest to LSTM, for the task of classifying a message as question or non-question. The comparison was done by training and optimizing the models on historic chat data from the Swedish insurance company Hedvig. Different types of word embedding were also tested, such as Word2vec and Bag of Words. The results demonstrated that LSTM achieved slightly higher scores than Random Forest, in terms of F1 and accuracy. The models’ performance were not significantly improved after optimization and it was also dependent on which corpus the models were trained on. An investigation of how a chatbot would affect Hedvig’s adoption rate was also conducted, mainly by reviewing previous studies about chatbots’ effects on user experience. The potential effects on the innovation’s five attributes, relative advantage, compatibility, complexity, trialability and observability were analyzed to answer the problem statement. The results showed that the adoption rate of Hedvig could be positively affected, by improving the first two attributes. The effects a chatbot would have on complexity, trialability and observability were however suggested to be negligible, if not negative. / Det vetenskapliga området språkteknologi har fått ökad uppmärksamhet den senaste tiden, men mindre fokus riktas på att jämföra modeller som skiljer sig i komplexitet. Den här kandidatuppsatsen jämför Random Forest med LSTM, genom att undersöka hur väl modellerna kan användas för att klassificera ett meddelande som fråga eller icke-fråga. Jämförelsen gjordes genom att träna och optimera modellerna på historisk chattdata från det svenska försäkringsbolaget Hedvig. Olika typer av word embedding, så som Word2vec och Bag of Words, testades också. Resultaten visade att LSTM uppnådde något högre F1 och accuracy än Random Forest. Modellernas prestanda förbättrades inte signifikant efter optimering och resultatet var också beroende av vilket korpus modellerna tränades på. En undersökning av hur en chattbot skulle påverka Hedvigs adoption rate genomfördes också, huvudsakligen genom att granska tidigare studier om chattbotars effekt på användarupplevelsen. De potentiella effekterna på en innovations fem attribut, relativ fördel, kompatibilitet, komplexitet, prövbarhet and observerbarhet analyserades för att kunna svara på frågeställningen. Resultaten visade att Hedvigs adoption rate kan påverkas positivt, genom att förbättra de två första attributen. Effekterna en chattbot skulle ha på komplexitet, prövbarhet och observerbarhet ansågs dock vara försumbar, om inte negativ.

The goal of this work is to expand the knowledge on the field of Venn Prediction employed with Survival Data. Standard Venn Predictors have been used with Random Forests and binary classification tasks. However, they have not been utilised to predict events with Survival Data nor in combination with Random Survival Forests. With the help of a Data Transformation, the survival task is transformed into several binary classification tasks. One key aspect of Venn Prediction are the categories. The standard number of categories is two, one for each class to predict. In this work, the usage of ten categories is explored and the performance differences between two and ten categories are investigated. Seven data sets are evaluated, and their results presented with two and ten categories. For the Brier Score and Reliability Score metrics, two categories offered the best results, while Quality performed better employing ten categories. Occasionally, the models are too optimistic. Venn Predictors rectify this performance and produce well-calibrated probabilities. / Målet med detta arbete är att utöka kunskapen om området för Venn Prediction som används med överlevnadsdata. Standard Venn Predictors har använts med slumpmässiga skogar och binära klassificeringsuppgifter. De har emellertid inte använts för att förutsäga händelser med överlevnadsdata eller i kombination med Random Survival Forests. Med hjälp av en datatransformation omvandlas överlevnadsprediktion till flera binära klassificeringsproblem. En viktig aspekt av Venn Prediction är kategorierna. Standardantalet kategorier är två, en för varje klass. I detta arbete undersöks användningen av tio kategorier och resultatskillnaderna mellan två och tio kategorier undersöks. Sju datamängder används i en utvärdering där resultaten presenteras för två och tio kategorier. För prestandamåtten Brier Score och Reliability Score gav två kategorier de bästa resultaten, medan för Quality presterade tio kategorier bättre. Ibland är modellerna för optimistiska. Venn Predictors korrigerar denna prestanda och producerar välkalibrerade sannolikheter.

With the availability of inexpensive hardware and software, digital imaging has become an important medium of communication in our daily lives. A huge amount of digital images are being collected and become available through the internet and stored in various fields such as personal image collections, medical imaging, digital arts etc. Therefore, it is important to make sure that images are stored, searched and accessed in an efficient manner. The use of bag of visual words (BOW) model for modelling images based on local invariant features computed at interest point locations has become a standard choice for many computer vision tasks. Based on this promising model, this thesis investigates three main problems: natural scene classification, annotation and retrieval. Given an image, the task is to design a system that can determine to which class that image belongs to (classification), what semantic concepts it contain (annotation) and what images are most similar to (retrieval). This thesis contributes to scene classification by proposing a weighting approach, named keypoints density-based weighting method (KDW), to control the fusion of colour information and bag of visual words on spatial pyramid layout in a unified framework. Different configurations of BOW, integrated visual vocabularies and multiple image descriptors are investigated and analyzed. The proposed approaches are extensively evaluated over three well-known scene classification datasets with 6, 8 and 15 scene categories using 10-fold cross validation. The second contribution in this thesis, the scene annotation task, is to explore whether the integrated visual vocabularies generated for scene classification can be used to model the local semantic information of natural scenes. In this direction, image annotation is considered as a classification problem where images are partitioned into 10x10 fixed grid and each block, represented by BOW and different image descriptors, is classified into one of predefined semantic classes. An image is then represented by counting the percentage of every semantic concept detected in the image. Experimental results on 6 scene categories demonstrate the effectiveness of the proposed approach. Finally, this thesis further explores, with an extensive experimental work, the use of different configurations of the BOW for natural scene retrieval. / Applied Science University in Jordan

В современном мире текстовая обработка и искусственный интеллект активно используются для автоматизации различных процессов, включая создание аннотаций для художественных произведений. Автоматическая генерация аннотаций помогает читателям быстро понять содержание книги и принять решение о её прочтении. В этой диссертации проводится исследование различных моделей генерации аннотаций, таких как Bag-of-Words (BoW), TF-IDF, Latent Dirichlet Allocation (LDA), Recurrent Neural Networks (RNNs), BERT (Bidirectional Encoder Representations from Transformers), T5 и PEGASUS. Эффективность этих моделей оценивается с помощью метрик BLEU Score, ROUGE Score, METEOR Score, F1 Score и CIDEr Score. Для тестирования моделей используется датасет, состоящий из книг в формате .docx. Результаты работы позволяют выявить наиболее эффективные методы автоматической генерации аннотаций и предлагают направления для дальнейшего совершенствования этих моделей. / In today's world, text processing and artificial intelligence are actively used to automate various processes, including the creation of annotations for fiction works. Automatic annotation generation helps readers quickly grasp the content of a book and decide whether to read it. This dissertation investigates various models for generating annotations, such as Bag-of-Words (BoW), TF-IDF, Latent Dirichlet Allocation (LDA), Recurrent Neural Networks (RNNs), BERT (Bidirectional Encoder Representations from Transformers), T5, and PEGASUS. The effectiveness of these models is evaluated using metrics such as BLEU Score, ROUGE Score, METEOR Score, F1 Score, and CIDEr Score. A dataset of books in .docx format is used to test the models. The results of the study identify the most effective methods for automatic annotation generation and suggest directions for further improvement of these models.

Thesis (MScEng (Mechanical and Mechatronic Engineering))--University of Stellenbosch, 2011. / Please refer to full text to view abstract. / Imported from http://etd.sun.ac.za. / np2011