Thèses sur le sujet « Document détection »
Créez une référence correcte selon les styles APA, MLA, Chicago, Harvard et plusieurs autres
Consultez les 50 meilleures thèses pour votre recherche sur le sujet « Document détection ».
À côté de chaque source dans la liste de références il y a un bouton « Ajouter à la bibliographie ». Cliquez sur ce bouton, et nous générerons automatiquement la référence bibliographique pour la source choisie selon votre style de citation préféré : APA, MLA, Harvard, Vancouver, Chicago, etc.
Vous pouvez aussi télécharger le texte intégral de la publication scolaire au format pdf et consulter son résumé en ligne lorsque ces informations sont inclues dans les métadonnées.
Parcourez les thèses sur diverses disciplines et organisez correctement votre bibliographie.
Poulain, d'Andecy Vincent. « Système à connaissance incrémentale pour la compréhension de document et la détection de fraude ». Thesis, La Rochelle, 2021. http://www.theses.fr/2021LAROS025.
Texte intégralThe Document Understanding is the Artificial Intelligence ability for machines to Read documents. In a global vision, it aims the understanding of the document function, the document class, and in a more local vision, it aims the understanding of some specific details like entities. The scientific challenge is to recognize more than 90% of the data. While the industrial challenge requires this performance with the least human effort to train the machine. This thesis defends that Incremental Learning methods can cope with both challenges. The proposals enable an efficient iterative training with very few document samples. For the classification task, we demonstrate (1) the continue learning of textual descriptors, (2) the benefit of the discourse sequence, (3) the benefit of integrating a Souvenir of few samples in the knowledge model. For the data extraction task, we demonstrate an iterative structural model, based on a star-graph representation, which is enhanced by the embedding of few a priori knowledges. Aware about economic and societal impacts because the document fraud, this thesis deals with this issue too. Our modest contribution is only to study the different fraud categories to open further research. This research work has been done in a non-classic framework, in conjunction of industrial activities for Yooz and collaborative research projects like the FEDER Securdoc project supported by la région Nouvelle Aquitaine, and the Labcom IDEAS supported by the ANR
Nguyen, Minh Tien. « Détection de textes générés automatiquement ». Thesis, Université Grenoble Alpes (ComUE), 2018. http://www.theses.fr/2018GREAM025/document.
Texte intégralAutomatically generated text has been used in numerous occasions with distinct intentions. It can simply go from generated comments in an online discussion to a much more mischievous task, such as manipulating bibliography information. So, this thesis first introduces different methods of generating free texts that resemble a certain topic and how those texts can be used. Therefore, we try to tackle with multiple research questions. The first question is how and what is the best method to detect a fully generated document.Then, we take it one step further to address the possibility of detecting a couple of sentences or a small paragraph of automatically generated text by proposing a new method to calculate sentences similarity using their grammatical structure. The last question is how to detect an automatically generated document without any samples, this is used to address the case of a new generator or a generator that it is impossible to collect samples from.This thesis also deals with the industrial aspect of development. A simple overview of a publishing workflow from a high-profile publisher is presented. From there, an analysis is carried out to be able to best incorporate our method of detection into the production workflow.In conclusion, this thesis has shed light on multiple important research questions about the possibility of detecting automatically generated texts in different setting. Besides the researching aspect, important engineering work in a real life industrial environment is also carried out to demonstrate that it is important to have real application along with hypothetical research
Martorell, Alexandre. « Détection à distance d’électroniques par l’intermodulation ». Thesis, Montpellier, 2018. http://www.theses.fr/2018MONTS019/document.
Texte intégralElectromagnetism, security and electronic warfare have been closely linked for decades. Their association gathers applications of radar surveillance, neutralization of electronic systems or detection of hidden electronics. Today, the multiplication of IEDs (Improvised Explosive Devices) both in theatres of operation and in urban environments leads to the need for their detection. The works of this thesis can enter into this theme and propose a new alternative that allows to highlight the presence of hidden RF receivers. The nonlinear radar is particularly suitable for detecting devices containing metals and (electronic) semiconductors. A popular technique is to transmit a single frequency f1 and receive the second harmonic generated by the target. Another less common technique consists of transmitting two frequencies, f1 and f2, and receiving intermodulation products of order 3 (2f1 - f2 and 2f2 - f1). An in-depth state of the art of nonlinear radar systems is made in a first chapter with a comparison of their characteristics. In a second chapter, an inductive test bench is developed to measure the reflected IM3 of an RF target. Thus analyses and orders of magnitude will be known helping the development of radar. In chapter 3, the IM3 radar demonstrator is developed. A wide range of RF systems, commercial and non-commercial, that may be found in operational environments are being tested. Their detection will validate the IM3 recovery technique. A new realistic IM3 radar link budget is implemented to estimate the actual radar detection range for different RF targets. In the last chapter the work focuses on the identification and classification of an RF target. The study focuses on the possibility of extracting all parameters to assist in a classification (hazard assessment) of RF receptors in an operational environment. The research work presented in this manuscript contributes to the improvement of hidden electronic detection techniques. A detection protocol was proposed describing the actions of the IM3 radar. It includes a frequency scan and then a power scan. The first tests were carried out on a walkie-talkie demonstrating the possibility of detecting its bandwidth via IM3 retransmission, at more than 2 m. The repeatability of the tests on an extended panel of RF receivers validates the detection protocol and the interest of the IM3 radar. An IM3 radar transmission power of 40 dBm, at an IM3 frequency of 400 MHz, can potentially detect a receiver at 80 m. Finally in a final exploratory work, we demonstrated that by observing the IM3 response reflected following a power scan the IM3 radar can add new identification criteria that discriminate the hidden receivers detected between them
Sharif, Md Haidar. « Détection et suivi d’événements de surveillance ». Thesis, Lille 1, 2010. http://www.theses.fr/2010LIL10034/document.
Texte intégralComputer vision algorithms have played a vital role in video surveillance systems to detect surveillance events for public safety and security. Even so, a common demerit among these systems is their unfitness to handle divers crowded scenes. In this thesis, we have developed algorithms which accommodate some of the challenges encountered in videos of crowded environments (e.g., airports, malls, sporting events) to a certain degree. We have adopted approaches by first performing a global-level motion analysis within each frame’s region of interest that provides the knowledge of crowd’s multi-modal behaviors in the form of complex spatiotemporal structures. These structures are then employed in the detection of unusual surveillance events occurred in the crowds. To conduct experiments, we have heavily relied on three thought-provoking datasets. The results reflect some unique global excellences of the approaches. We have also developed a pseudo Euclidian distance. To show its usage, a methodology based on it has been employed in the detection of various usual surveillance events from theTRECVID2008. Some results report the robustness of the methodology, while the rest gives evidence of the difficulty of the problem at hand. Big challenges include, but are not limited to, massive population flow, heavy occlusion, reflection, shadow, fluctuation, varying target sizes, etc. Notwithstanding, we have got much useful insights and experience to the practical problems. In addition, the thesis explores an individual target tracking algorithm within miscellaneous crowded scenes. Video sequences from the PETS2009 Benchmark data have been used to evaluate its performance. Viewing its pros and cons, the algorithm is still highly accurate and its sensitivity to the effects of diversity in noise and lighting, which ascertains its high-quality performance on disappearances, targets moving in and out of the shadow, and flashes of light
Donatin, Emilie. « Détection des bactéries entéropathogènes : approche polyphasique ». Thesis, Aix-Marseille, 2012. http://www.theses.fr/2012AIXM5034/document.
Texte intégralThe human body is a collection of microflora where cohabit bacteria, archaea, viruses and eukaryotes. These complex ecosystems are called microbiota. Among these is the intestinal microbiota that counts 1011 to 1014 bacteria/g of stool. Changes in the intestinal flora can cause of pathologies such as infectious diarrhea. This is a real public health problem since about 2.16 million deaths are related to this disease each year. Enteric viruses play a preponderant role but bacterial infections are also important. The diagnosis of bacterial infections is complicated because the intestinal microbiota includes 75% non-cultivable species. In addition, there is not really a list of bacteria could be responsible for infectious diarrhea. We therefore decided to study the intestinal microbiota in normal stool and also pathological stools by a polyphasic approach combining a preliminary step of diarrheal stools concentration by lyophilization, with cultivation techniques and molecular biology methods. We developed a new technology for the detection of enteropathogens by hybridization on DNA microarray for the detection of bacteria and enteric viruses (DNA) in the presence of a control archaea. Our tool allows multiplexed diagnostic of infectious diarrhea since we correctly identified an adenovirus and Campylobacter jejuni present in a same sample. This is the first DNA microarray for multiplex detection of bacteria and viruses (DNA) enteropathogens. An improvement of our protocol for nucleic acid extraction is proposed to allow the detection of RNA viruses such as rotavirus and calicivirus which are currently dominant
Le, Danvic Chrystelle. « Détection de l’oestrus chez les bovins ». Thesis, Lille 1, 2009. http://www.theses.fr/2009LIL10185/document.
Texte intégralEstrus detection is the critical stage in livestock reproduction. Efficiency of artificial insemination, the major method of reproduction used in cattle, depends on the accurate detection of this short female stage. Currently used detection methods are quite expensive and insecure. This work aims to the elucidation of chemical communication involved in estrus detection by the male. Indeed, male can detect heats by olfactory cues emitted in urine of female. Identification of such chemical cues and their associated binding proteins in estrus behaviour will permit to develop new biotechnological tools as biosensors for estrus detection. To characterize estrus specific molecules, urine from 30 cows was collected at specific stages of the estrus cycle and their chemical composition has been assayed by GC/MS analysis. No systematic estrus specific compounds have been characterized, but we were able to characterize a few compounds as pre-estrus and/or estrus specific in some animals. Identified compounds remain to be tested for their biological activity and ability to elicit sexual behaviour to confirm their implication in estrus detection. No protein binding these molecules could be identified in urine. The urinary serum albumin, described in female elephant urine as a pheromone carrier protein, was characterized. The study of protein in the olfactory area led us to identify several isoforms of bovine Odorant Binding Protein (bOBP), in male and female olfactory tissues. Isoforms can be distinguished by their post-translational modifications. Biochemical experiments showed phosphorylation on one bOBP variant and a N-glycan is present at the N-terminus of the protein. Existence of various bOBP isoforms suggests function specificity, in particular in their binding properties. Spectroscopic fluorescence experiments were performed to analyze the structure-function relationships between the various bOBP isoforms and odorant molecules to ascertain this hypothesis
Girardeau, Vadim. « Imagerie photo-acoustique à détection optique ». Thesis, Université Grenoble Alpes (ComUE), 2018. http://www.theses.fr/2018GREAT049/document.
Texte intégralIn the context of an aging population, it is essential to develop clinical diagnostic tools that are accurate, reliable, inexpensive and easy to implement. During this thesis I particularly sought to perform a dynamic mapping of blood vessels in order to detect both cancers and cardiovascular diseases, two of the most deadly diseases. In order to have an effective diagnosis, it is necessary to image in depth with the best possible spatial and temporal resolution. In chapter 0 I explain the challenges of medical imaging on micro-vascularization by analyzing the advantages and disadvantages of several types of medical imaging. In chapter 1 I develop in detail the photo-acoustic imaging, which proved to be the most appropriate for our application. It has the advantage of optical contrast and acoustic resolution. In particular, I use frequency photo-acoustics, which is inexpensive and space-saving, and can therefore be easily integrated in the hospital world compared to "traditional" photo-acoustic imaging. I validate this part on in-vivo experimental results on mouse ears. In chapter 2 I tried to detect the photo-acoustic signal in an optical way which has the advantage of being contactless and therefore without any problem of clutter between "optical excitation" and "acoustic detection". I develop the signal processing necessary to detect an acoustic wave, i.e. vibrations, using an interferometer. Then I present in chapter 3 a particular interferometer developed in the laboratory: the Laser Optical Feedback Imaging (LOFI). This interferometer allows to be limited to photon noise even with a low intensity thus it makes it possible to detect vibrations of small amplitudes even on surfaces with a low reflecting index like the skin in accordance with medical standards. In chapter 4 I validate the detection of the photo-acoustic signal with our optical detection. Finally in chapter 5 I show with simulations an innovative full field imaging technique that would allow faster detection of a spectrally rich photo-acoustic signal
Bouvier, des Noes Mathieu. « Détection itérative des séquences pseudo-aléatoires ». Thesis, Université Grenoble Alpes (ComUE), 2015. http://www.theses.fr/2015GREAT068/document.
Texte intégralPseudo-random binary sequences are very common in wireless transmission systems and ciphering mechanisms. More specifically, they are used in direct sequence spread spectrum transmission systems like UMTS or GPS, or to construct preamble sequences for synchronization and channel estimation purpose like in LTE. It is always required to synchronize the receiver with the transmitted sequence. The usual way consists in correlating the received signal with a replica of the sequence. If the correlation exceeds a predefined threshold, the synchronization is declared valid.This thesis addresses a different approach: the binary sequence is detected with a forward error correction decoding algorithm. This allows for instance to detect very long sequences.In this thesis, we show that decoding a pseudo-random sequence is a problematic of the kind ‘detect and decode'. The decoder detects the presence of the transmitted sequence and simultaneously estimates its initial state. In conventional detection theory, this corresponds to a GLRT detector that uses a decoder to estimate the unknown parameter which is the transmitted sequence. For pseudo-random sequences, the decoder implements an iterative message-passing algorithm. It uses a parity check matrix to define the decoding graph on which the algorithm applies. Each parity check equation has a weight t, corresponding to the number of variables in the equation.Parity check equations are thus an essential component of the decoder. The decoding procedure is known to be sensitive to the weight t of the parity check equations. For m-sequences, the number of parity check equations is already known. It is given by the number of codewords of weight t of the corresponding Hamming dual code. For Gold sequences, the number of parity check equations of weight t = 3 and 4 has already been evaluated by Kasami. In this thesis we provide an analytical expression for the number of parity check equations of weight t = 5 when the degree of the generator polynomial r is odd. Knowing this number is important because there is no parity check equation of weight t < 5 when r is odd. This enumeration is also used to provide an estimation of the least degree of parity check equations of weight t.We have then addressed the problem of selecting the parity check equations used by the decoder. We observed the probability of false alarm is very sensitive to this selection. It is explained by the presence or absence of absorbing sets which block the convergence of the decoder when it is fed only with noise. These sets are known to be responsible for error floor of LDPC codes. We give a method to identify these sets according to the parity check equations used by the decoder. The probability of false alarm can increase dramatically if these absorbing sets are destroyed. Then we propose an algorithm for selecting these parity check equations. It relies on the minimization of the number of cycles of length 6 and 8. Simulation show that the algorithm allows to improve significantly the probability of false alarm and the average acquisition time.Eventually, we propose 2 algorithms for the detection of the scrambling codes used in the uplink of UMTS-FDD and CDMA2000 systems. They highlights a new vulnerability of DSSS transmission systems. It is now conceivable to detect these transmission if the sequence's generator is known
Lesmanne, Emeline. « Approche photonique de la détection multispectrale ». Thesis, Université Grenoble Alpes (ComUE), 2016. http://www.theses.fr/2016GREAY097/document.
Texte intégralThis thesis investigates frequency selective surfaces as filters for infrared multispectral imaging. Frequency selective surfaces for the transmission of light are made of a metallic layer etched with periodic resonant apertures. The resonance wavelength and the filter's parameters (transmission, full width at half maximum) depend on lateral geometrical parameters, facilitating the design and integration of filters on the focal plane array of the detector. A theoretical model using the temporal coupled mode theory will be used to highlight key parameters of the physics of these structures at infrared wavelengths (λ=1-12 µm), where metals are absorbing. Radiative and absorptive loss rates will be calculated and commented. FDTD (Finite Difference Time Domain) simulations will be made and compared with the model. Furthermore, optical stacks taking into account technological constraints will be simulated and analyzed and will allow us to define an optimal optical stack, considering geometrical and material parameters (choice of optical indexes). The filters' angular tolerance will be considered
Mansour, Elio. « Détection d’évènements dans des environnements connectés ». Thesis, Pau, 2019. http://www.theses.fr/2019PAUU3017/document.
Texte intégralThe rising interest in smart connected environments (e.g., smart buildings, cities, factories) and theevolution of sensors, data management/communication technologies have paved the way forinteresting and useful applications that help users in their every day tasks (e.g. increasing comfort,reducing energy consumption). However, various improvements are still required. For instance, howto enhance the representation of such complex, dynamic, and heterogeneous environments.Moreover, how to facilitate the interaction between users and their connected environments, and howto provide tools for environment monitoring and management.In this thesis, we focus on four main challenges: (i) representing a diverse set of components andelements related to the environment and its sensor network; (ii) providing a query language thathandles user/connected environment interactions (e.g., environment definition, data management,event definition); (iii) coping with the dynamicity of the environment and its evolution over time; and(iv) proposing a generic event detection mechanism for improved environment monitoring.To do so, we first present an ontology-based data model that represents hybrid environments/sensornetworks. Thus covering diverse sensors (e.g., static, mobile), environments (e.g., infrastructures,devices), and data (e.g., scalar, multimedia). Then, we introduce a query language that one might usefor various tasks (e.g., defining the connected environment, information retrieval, event definition,data management). Furthermore, to keep up with the environment changes we provide a queryoptimizer that allows the submitted queries to cope with the dynamicity of the environment prior totheir execution. Finally, we propose an event detection core that takes event definitions as input anddetects the targeted events.We group the aforementioned modules in one global framework for event detection in connectedenvironments. Our proposal is generic, extensible, and could be used with different connectedenvironments such as buildings, cities. .
David, Ciprian Petru. « Détection d'hétérogénéités linéaires dans les textures directionnelles : application à la détection de failles en sismique de réflexion ». Thesis, Bordeaux 1, 2008. http://www.theses.fr/2008BOR13714/document.
Texte intégralLinear disparity detection in directional textures – application to fault detection in seismic images: the developed approaches deal with the particular family of directional textures. In the first chapter the notion of texture and the concept of contour are introduced. Also, a detailed presentation of the application concerning seismic imagery is presented in the first chapter. The object of the second chapter is the analysis of the different contributions concerning edge and contour detection in textured images found in the literature: the approaches used in the field of geophysics and the approaches proposed by the image processing community. The third chapter regroups our contributions: a geometric criterion based approach, a recursive robust extension of the geometric approach and a 3D recursive robust extension combining a 2D measure and a 3D diffusion technique. Apart the qualitative comparisons, these contributions are validated by a quantitative analysis in comparison with the existing methods
Aberkane, Arezki. « Détection de contours dans les images CFA ». Thesis, Lille 1, 2017. http://www.theses.fr/2017LIL10222/document.
Texte intégralThis thesis is devoted to edge detection from the raw image acquired by single-sensor cameras. Such cameras are fitted with Bayer Color Filter Array (CFA, generally Bayer one) and deliver raw CFA image, in which each pixel is characterized by only one out of the three colour components (red, green, or blue). A demosaicing procedure is necessary to estimate the other two missing colour components at each pixel, so as to obtain a colour image. This however produces artefacts that may affect the performance of low-level processing tasks applied to such estimated images. We propose to avoid demosaicing to compute partial derivatives for edge detection. Simple differentiation kernels, Deriche filters or shifted Deriche filters can be used either in a vector or a scalar approache. The vector approach computes partial derivatives for the three channels and the scalar approach first estimates a luminance image, then computes derivatives. The partial CFA derivatives are then used to compute Di Zenzo gradient for edge detection. We assess the performance of our methods on a large dataset of synthetic images with available edge ground truth. We show that CFA-based approaches may provide as accurate edge detection results as colour vector-based ones at much reduced computation cost
Palisse, Aurélien. « Analyse et détection de logiciels de rançon ». Thesis, Rennes 1, 2019. http://www.theses.fr/2019REN1S003/document.
Texte intégralThis phD thesis takes a look at ransomware, presents an autonomous malware analysis platform and proposes countermeasures against these types of attacks. Our countermeasures are real-time and are deployed on a machine (i.e., end-hosts). In 2013, the ransomware become a hot subject of discussion again, before becoming one of the biggest cyberthreats beginning of 2015. A detailed state of the art for existing countermeasures is included in this thesis. This state of the art will help evaluate the contribution of this thesis in regards to the existing current publications. We will also present an autonomous malware analysis platform composed of bare-metal machines. Our aim is to avoid altering the behaviour of analysed samples. A first countermeasure based on the use of a cryptographic library is proposed, however it can easily be bypassed. It is why we propose a second generic and agnostic countermeasure. This time, compromission indicators are used to analyse the behaviour of process on the file system. We explain how we configured this countermeasure in an empiric way to make it useable and effective. One of the challenge of this thesis is to collate performance, detection rate and a small amount of false positive. To finish, results from a user experience are presented. This experience analyses the user's behaviour when faced with a threat. In the final part, I propose ways to enhance our contributions but also other avenues that could be explored
Caubet, Fabien. « Détection d’un objet immergé dans un fluide ». Thesis, Pau, 2012. http://www.theses.fr/2012PAUU3006/document.
Texte intégralThis dissertation takes place in the mathematic field called shape optimization. More precisely, we focus on a detecting inverse problem using shape calculus and asymptotic analysis. The aim is to localize an object immersed in a viscous, incompressible and stationary fluid. This work was motivated by the following main questions:– can we localize an obstacle immersed in a fluid from a boundary measurement?– can we reconstruct numerically this object, i.e. be close to its localization and its shape, from this measure?– can we know how many objects are included in the fluid using this measure?The results are described in the five chapters of the thesis:– the first one gives a mathematical framework in order to prove the existence of the shape derivatives oforder one and two in the frame of the detection of inclusions;– the second one analyzes the detection problem using geometric shape optimization: an identifiabilityresult is proved, the shape gradient of several shape functionals is characterized and the instability of thisinverse problem is proved;– the chapter 3 uses our theoretical results in order to reconstruct numerically some objets immersed in a fluid using a shape gradient algorithm;– the fourth chapter analyzes the detection of small inclusions in a fluid using the topological shape optimization : the topological gradient of a Kohn-Vogelius shape functional is characterized;– the last chapter uses this theoretical expression in order to determine numerically the number and the location of some small obstacles immersed in a fluid using a topological gradient algorithm
Wang, Yan. « Détection des changements à partir de photographies ». Thesis, Toulouse 3, 2016. http://www.theses.fr/2016TOU30069/document.
Texte intégralThis work deals with change detection from chronological series of photographs acquired from the ground. This context of consecutive images comparison is the one encountered in the field of integrated geography where photographic landscape observatories are widely used. These tools for analysis and decision-making consist of databases of photographic images obtained by strictly rephotographing the same scene at regular time intervals. With a large number of images, the human analysis is tedious and inaccurate. So a tool for automatically comparing pairs of landscape photographs in order to highlight changes would be a great help for exploiting photographic landscape observatories. Obviously, lighting variations, seasonality, time of day induce completely different images at the pixel level. Our goal is to design a system which would be robust to these insignificant changes and able to detect relevant changes of the scene. Numerous studies have been conducted on change detection from satellite images. But the utilization of classic digital cameras from the ground raise some specific problems like the limitation of the spectral band number and the strong variation of the depth in a same image which induces various appearance of the same object categories depending on their position in the scene. In the first part of our work, we investigate the track of automatic change detection. We propose a method lying on the registration and the over-segmentation of the images into superpixels. Then we describe each superpixel by its texture using texton histogram and its gray-level mean. A distance measure, such as Mahalanobis distance, allows to compare corresponding superpixels between two images acquired at different dates. We evaluate the performance of the proposed approach on images taken from the photographic landscape observatory produced during the construction of the French A89 highway. Among the image segmentation methods we have tested for superpixel extraction, our experiments show the relatively good behavior of Achanta segmentation method. The relevance of a change is strongly related to the intended application, we thus investigate a second track involving a user intervention. We propose an interactive change detection method based on a learning step. In order to detect changes between two images, the user designates with a selection tool some samples consisting of pixel sets in "changed" and "unchanged" areas. Each corresponding pixel pair, i.e., located at the same coordinates in the two images, is described by a 16-dimensional feature vector mainly calculated from the dissimilarity image. The latter is computed by measuring, for each corresponding pixel pair, the dissimilarity of the gray-levels of the neighbors of the two pixels. Samples selected by the user are used as learning data to train a classifier. Among the classification methods we have tried, experimental results indicate that random forests give the better results for the tested image series
Katz, Gil. « Détection binaire distribuée sous contraintes de communication ». Thesis, Université Paris-Saclay (ComUE), 2017. http://www.theses.fr/2017SACLC001/document.
Texte intégralIn recents years, interest has been growing in research of different autonomous systems. From the self-dring car to the Internet of Things (IoT), it is clear that the ability of automated systems to make autonomous decisions in a timely manner is crucial in the 21st century. These systems will often operate under stricts constains over their resources. In this thesis, an information-theoric approach is taken to this problem, in hope that a fundamental understanding of the limitations and perspectives of such systems can help future engineers in designing them.Throughout this thesis, collaborative distributed binary decision problems are considered. Two statisticians are required to declare the correct probability measure of two jointly distributed memoryless process, denoted by $vct{X}^n=(X_1,dots,X_n)$ and $vct{Y}^n=(Y_1,dots,Y_n)$, out of two possible probability measures on finite alphabets, namely $P_{XY}$ and $P_{bar{X}bar{Y}}$. The marginal samples given by $vct{X}^n$ and $vct{Y}^n$ are assumed to be available at different locations.The statisticians are allowed to exchange limited amounts of data over a perfect channel with a maximum-rate constraint. Throughout the thesis, the nature of communication varies. First, only unidirectional communication is allowed. Using its own observations, the receiver of this communication is required to first identify the legitimacy of its sender by declaring the joint distribution of the process, and then depending on such authentication it generates an adequate reconstruction of the observations satisfying an average per-letter distortion. Bidirectional communication is subsequently considered, in a scenario that allows interactive communication between the participants
Favard, Alexandre. « Multicapteurs intégrés pour la détection des BTEX ». Thesis, Aix-Marseille, 2018. http://www.theses.fr/2018AIXM0123/document.
Texte intégralOutdoor air quality is subjected to the law LAURE since 1996. In 2008, the european directive 2008/50/EC introduced measurement requirements and thresholds that should not be exceeded for certain pollutants on a european scale. According to several toxicological and epidemiological studies, air pollution causes respiratory failure, asthma, cardiovascular diseases and cancers. In Europe, air pollution is responsible for more than 300 000 early deaths a year.Volatile organic compounds (VOCs), particularly benzene, toluene, ethylbenzene and xylenes (BTEX compounds) are proven pollutants and play a major role in the degradation of indoor and outdoor air quality. This thesis is dedicated to the development of a metal oxide based multi-gas sensor for the detection of traces of BTEX within the framework of the SMARTY project (SMart AiR qualiTY). A complete electrical characterization system was designed and implemented for the detection of sub-ppm concentrations of BTEX.Based on the state-of-art, several materials were selected (WO3, ZnO, SnO2). The electrical characterizations of the selected sensitive layers were carried out under dry air and under different humidity levels in the presence of BTEX and interfering gases (NO2, CO2). Tungsten oxide (WO3) exhibits the best performance in the presence of moisture and is chosen for the technology transfer that accompanies the new patented AMU transducers. The WO3-based multi-sensor has a lower limit of detection (LOD) of 1 ppb at 50% relative humidity and effectively detects and quantifies BTEX
Derbas, Nadia. « Contributions à la détection de concepts et d'événements dans les documents vidéos ». Thesis, Grenoble, 2014. http://www.theses.fr/2014GRENM035/document.
Texte intégralA consequence of the rise of digital technology is that the quantity of available collections of multimedia documents is permanently and strongly increasing. The indexing of these documents became both very costly and impossible to do manually. In order to be able to analyze, classify and search multimedia documents, indexing systems have been defined. However, most of these systems suffer quality or practicability issues. Their performance is limited and depends on the data volume and data variability. Indexing systems analyze multimedia documents, looking for static concepts (bicycle, chair,...), or events (wedding, protest,...). Therefore, the variability in shapes, positions, lighting or orientation of objects hinders the process. Another aspect is that systems must be scalable. They should be able to handle big data while using reasonable amount of computing time and memory.The aim of this thesis is to improve the general performance of content-based multimedia indexing systems. Four main contributions are brought in this thesis for improving different stages of the indexing process. The first one is an "early-early fusion method" that merges different information sources in order to extract their deep correlations. This method is used for violent scenes detection in movies. The second contribution is a weakly supervised method for basic concept (objects) localization in images. This can be used afterwards as a new descriptor to help detecting complex concepts (events). The third contribution tackles the noise reduction problem on ambiguously annotated data. Two methods are proposed: a shot annotation generator, and a shot weighing method. The last contribution is a generic descriptor optimization method, based on PCA and non-linear transforms.These four contributions are tested and evaluated using reference data collections, including TRECVid and MediaEval. These contributions helped our submissions achieving very good rankings in those evaluation campaigns
Zitzmann, Cathel. « Détection statistique d'information cachée dans des images naturelles ». Thesis, Troyes, 2013. http://www.theses.fr/2013TROY0012/document.
Texte intégralThe need of secure communication is not something new: from ancient, methods exist to conceal communication. Cryptography helped make unintelligible message using encryption, steganography can hide the fact that a message is exchanged.This thesis is part of the project "Hidden Information Research" funded by the National Research Agency, Troyes University of Technology worked on the mathematical modeling of a natural image and creating detectors of hidden information in digital pictures.This thesis proposes to study the steganalysis in natural images in terms of parametric statistical decision. In JPEG images, a detector based on the modeling of quantized DCT coefficients is proposed and calculations of probabilities of the detector are established theoretically. In addition, a study of the number of shrinkage occurring during embedding by F3 and F4 algorithms is proposed. Finally, for the uncompressed images, the proposed tests are optimal under certain constraints, a difficulty overcome is the data quantization
Koucham, Oualid. « Détection d'intrusions pour les systèmes de contrôle industriels ». Thesis, Université Grenoble Alpes (ComUE), 2018. http://www.theses.fr/2018GREAT090/document.
Texte intégralThe objective of this thesis is to develop intrusion detection and alert correlation techniques geared towards industrial control systems (ICS). Our interest is driven by the recent surge in cybersecurity incidents targeting ICS, and the necessity to detect targeted attacks which induce incorrect behavior at the level of the physical process.In the first part of this work, we develop an approach to automatically infer specifications over the sequential behavior of ICS. In particular, we rely on specification language formalisms such as linear temporal logic (LTL) and metric temporal logic (MTL) to express temporal properties over the state of the actuators and sensors. We develop an algorithm to automatically infer specifications from a set of specification patterns covering the most recurring properties. In particular, our approach aims at reducing the number of redundant and unfalsifiable properties generated by the existing approaches. To do so, we add a pre-selection stage which allows to restrict the search for valid properties over non redundant portions of the execution traces. We evaluate our approach on a complex physical process steered by several controllers under process oriented attacks. Our results show that a significant reduction in the number of inferred properties is possible while achieving high detection rates.In the second part of this work, we attempt to combine the physical domain intrusion detection approach developed in the first part with more classical cyber domain intrusion detection approaches. In particular, we develop an alert correlation approach which takes into account some specificities of ICS. First, we explore an alert enrichment approach that allows to map physical domain alerts into the cyber domain. This is motivated by the observation that alertscoming from different domains are characterized by heterogeneous attributes which makes any direct comparison of the alerts difficult. Instead, we enrich the physical domain alerts with cyber domain attributes given knowledge about the protocols supported by the controllers and the memory mapping of process variables within the controllers.In this work, we also explore ICS-specific alert selection policies. An alert selection policy defines which alerts will be selected for comparison by the correlator. Classical approaches often rely on sliding, fixed size, temporal windows as a basis for their selection policy. Instead, we argue that given the complex interdependencies between physical subprocesses, agreeing on analert window size is challenging. Instead, we adopt selection policies that adapt to the state of the physical process by dynamically adjusting the size of the alert windows given the state of the subprocesses within the physical process. Our evaluation results show that our correlator achieves better correlation metrics in comparison with classical temporal based approaches
Farhat, Ahmad. « Détection, localisation et estimation de défauts : application véhicule ». Thesis, Université Grenoble Alpes (ComUE), 2016. http://www.theses.fr/2016GREAT056/document.
Texte intégralModern vehicles are increasingly equipped with new mechanisms to improve safety, comfort and ecological impact. These active systems employ sensors, actuators and automatic control systems. However, in case of failure of one these components, the consequences for the vehicle and the passengers safety could be dramatic. In order to ensure a higher level of reliability within on board diagnosis, new methodologies for sensor or actuator fault detection, location and estimation are proposed. These model based approaches are extended for robust synthesis for switched uncertain systems. In addition, a method for detecting critical stability situation is presented. The validation of the different methods is illustrated with simulations using CarSim, and application on real vehicle data within the INOVE project
Dadouchi, Florian. « Détection robuste de signaux acoustiques de mammifères marins ». Thesis, Grenoble, 2014. http://www.theses.fr/2014GRENU059/document.
Texte intégralThe oceans experience heavy anthropogenic pressure due to overfishing, physico-chemical pollution, and noise radiated by industrial and military activities. This work focuses on the use of passive acoustic monitoring of the oceans, as a tool to understand the impact of radiated noise on marine ecosystems, and particularly on marine mammals. This work tackles the task of detection of acoustical signals of marine mammals using the spectrogram. This task is uneasy for two reasons : 1. the ocean noise structure is complex (non-stationary and colored) and 2. the signals of interest are unknown and also shows a complex structure (non-stationary narrow band and/or impulsive). The problem therefore must be solved locally without making a priori hypothesis on the signal. Statistical detectors only based on the local analysis of the noise spectrogram coefficients are available, making them suitable for this problem. However, these detectors suffer two disadvantages : 1. the trade-offs false alarm probability/ detection probability that are available for low signal tonoise ratio are not satisfactory and 2. the separation between narrow-band and impulsive signals is not possible. This work brings some answers to these problems.The main contribution of this work is to formulate a binary hypothesis test taking explicitly in account the spatial organization of time-frequency peaks. We introduce the False Alarm Density Analysis (FADA) framework that efficiently discriminates time-frequency regions hosting signal from the ones hosting noise only. In particular the number of false alarms in regions of the binary spectrogram is first modeled by a binomial distribution, and then by a correlated binomial distribution to take in account the spectrogram redundancy. The binary hypothesis test is solved using a Neyman-Pearson criterion.We demonstrate the relevance of this approach on simulated data and validate the FADA detector on a wide variety of real signals. In particular we show the capability of the proposed method to efficiently detect signals in highly impulsive environment
Le, Bot Olivier. « Détection, localisation, caractérisation de transitoires acoustiques sous-marins ». Thesis, Grenoble, 2014. http://www.theses.fr/2014GRENU028/document.
Texte intégralThe underwater environment is insonified by a wide variety of acoustic sourcesthat can be monitored by autonomous passive acoustic recorders. A large number of the recordedsounds are transient signals (short-finite duration signals), among which the pulse signals that westudy in this thesis. Pulse signals have specific properties, such as a very short duration (<1ms), fewoscillations, a high directivity, which make them difficult to study by classical signal processing tools(Fourier transform, autocorrelation).In the first part of this study, we develop a method to detect sound sources emitting rhythmic pulsetrains (dolphins, sperm whales, beluga whales). This detector uses only the time of arrival of pulses atthe hydrophone to perform a rhythm analysis based on a complex autocorrelation and a time-rhythmrepresentation. This allows : i) to detect rhythmic pulse trains, ii) to know the beginning and endingtimes of pulse trains, iii) to know the value of the rhythm.In the second part of this thesis, we study the potential of a method called Recurrence Plot Analysis tocharacterize waveforms of pulse signals. After a general presentation of this method we develop threesignal processing architectures based on it, to perform the following tasks : i) transient detection, ii)transient characterization and pattern recognition, iii) estimation of time difference of arrival of thetransient on two hydrophones.All the methods developped in this thesis are validated on simulated and real data recorded at sea
Hadjerci, Oussama. « Détection automatique du nerf dans les images échographiques ». Thesis, Orléans, 2017. http://www.theses.fr/2017ORLE2006/document.
Texte intégralRegional anesthesia presents an interesting alternative or complementary act to general anesthesia in many surgical procedures. It reduces pain scores, improves postoperative mobility and facilitates earlier hospital discharge. Ultrasound-Guided Regional Anesthesia (UGRA) has been gaining importance in the last few years, offering numerous advantages over alternative methods of nerve localization (neurostimulation or paraesthesia). However, nerve detection is one of the most difficult tasks that anesthetists can encounter in the UGRA procedure. The context of the present work is to provide practitioners with a method to facilitate and secure the practice of UGRA. However, automatic detection and segmentation in ultrasound images is still a challenging problem in many medical applications. This work addresses two main issues. The first one, we propose an algorithm for nerve detection and segmentation in ultrasound images, this method is composed of a pre-processing, texture analysis and machine learning steps. In this part of work, we explore two new approaches ; one to characterize the nerve and the second for selecting the minimum redundant and maximum relevant features. The second one, we studied the nerve detection in consecutive ultrasound frames. We have demonstrated that the development of an algorithm based on the temporal coherence of the position, the shape and the confidence measure of the classification, allows to generate a robust segmentation. In this work, we also propose a new model of shape based on a set of intervals landmarks able to adapt to the nerve shape under a morphological variations
Milutinovic, Miléna. « Immobilisation d'hydrogel redox pour la détection par électrochimiluminescence ». Thesis, Bordeaux 1, 2011. http://www.theses.fr/2011BOR14411/document.
Texte intégralThe main goal of this thesis was to study electrogenerated chemiluminescence (ECL) and its application in development of new analytical techniques. Due to its high sensitivity, ECL presents a powerful method for applications in clinical diagnostic and environmental chemistry (presence of contaminants in water or food). The immobilisation of an ECL luminophore is usually performed on a solid phase. This step is an essential point to obtain a technique for fast, simple, flexible and effective immobilisation of ECL luminophores with possibility of applications in various configurations. The first part of this work presents the optimisation of the electrochemical deposition of a ruthenium metallopolymer and its application in enzymatic detection. A redox hydrogel film with micrometric thickness was prepared using cyclic voltammetry. This immobilised hydrogel allows the detection of model substrate (glucose) using enzyme glucose dehydrogenase. The second part of this thesis is focused on the development of a new photodeposition method for the ECL polymer immobilisation. This method allows region-selective immobilisation of active centres. Using photolithographic methods, the figures from the mask are projected on the electrode surface. This allows the formation of micrometric spots which size, shape and thickness is modulated. Electrochemical properties of obtained nanometric films are comparable with those of electrodeposited films. Also, ECL spectra recorded with both immobilisation strategies are identical. It shows that the ECL excited state is the same. The obtained photopatterns were imaged using ECL. The development of such process presents an alternative for realisation of different ECL spot arrays and allows multiplexed detection by ECL imaging. In the third part of this work we have associated spectroelectrochemistry and ECL imaging to study the ECL mechanisms at the level of a single microbead, functionalised with ruthenium complex. Combining fluorescence microscopy and ECL imaging, the distribution of electroactive and ECL sites have been highlighted. From this study we can clarify the mechanism that leads to ECL emission at the level of functionalised beads
Peyrard, Lisa. « Cyclotrivératrylènes fluorescents pour la détection d’ammoniums d’intérêt biologique ». Thesis, Bordeaux 1, 2012. http://www.theses.fr/2012BOR14700/document.
Texte intégralThe development of fluorescent probes for the ex vivo detection of biological phenomena (neuronal communication, for example) presents a growing interest. In this context, fluorescent cyclotriveratrylene (CTV) derivatives have been developed, as these cavitands are known to complex small organic molecules. The strategy used to get interesting spectroscopic properties, was to introduce conjugated electron-donating and electron-withdrawing groups on each aromatic unit of the CTV (leading to “push-pull” systems). To improve the spectroscopic properties of the probes, the conjugation between the electron-donating and the electron-withdrawing groups was extended. Hence, the Sonogashira organometalic coupling reaction and the copper catalyzed cycloaddition directly on CTV skeleton were used. The spectroscopic characterization of the new fluorescent probes synthesized was done, in organic solvent but also in physiological conditions when the solubility permits it. The recognition studies for biological ammoniums, such as neurotransmitters (acetylcholine, dopamine, and derivatives) were performed in physiological conditions by spectrofluorimetry but also other methods (like MNR or micro-calorimetry). The formation of molecular assemblies was also observed (by dynamic light scattering and transmission electron microscopy) during this work with some of the probes
Balique, Fanny. « Détection et impact potentiel des tobamovirus chez l'homme ». Thesis, Aix-Marseille, 2013. http://www.theses.fr/2013AIXM5034/document.
Texte intégralA current paradigm in virology is that plant viruses cannot affect vertebrates and are not pathogenic for humans. However, several recent findings challenge this paradigm. The aim of this thesis was to investigate whether tobamoviruses may be pathogenic in humans. This manuscript contains: 1) A review of the literature of the reasons why plant viruses may cross the border from plants to vertebrates and the possible impact of plant viruses in animals including humans. 2) A study on exposure to tobamovirus and possible effects of these viruses for humans. In the Marseille geographical area, we showed the presence PMMoV in 7.2% of stools from patients tested and in 57% of food products containing hot peppers. In addition, the presence of PMMoV in the stools of patients could be significantly correlated with clinical symptoms. Then, we showed that infectious TMV was present in the tobacco of all cigarettes tested and TMV RNA was detected in 45% of smokers' saliva tested. 3) In vivo study: mice intra-tracheal inoculation with TMV. The results showed the persistence until 14 days of viruses in the lung tissue and in lung alveolar macrophages of mice. In addition, we detected TMV in the cytoplasm of murine macrophages up to 15 days after inoculation. 4) A study of TMV transmission to humans through cigarette smoking. TMV does not seem to be vehicled by cigarette smoke to smoker's saliva but by direct contact with infected tobacco. Our results suggest that the boundary between plant viruses and animal viruses is not as strict as it is commonly accepted and prompt to reassess the potential pathogenicity of these plant viruses for humans
Frabolot, Ferdinand. « Optimisation de forme avec détection automatique de paramètres ». Thesis, Compiègne, 2015. http://www.theses.fr/2015COMP2182/document.
Texte intégralThe objective of this thesis work is to be able to completely integrate shape optimization of car inner hood stiffeners in a complex industrial process, in order to fully optimize the shape and distribution of the stiffeners in a multi-objective approach (or even multi-disciplinary) of a 3D surfacic structure. To this end, we established, at the outset, an insight of the state-of-the-art in shape optimization of structures by classifying the different shape parametrizations in three distinct categories : geometry-based methods (a shape parametrization such as a CAD model), grid-based methods (such as topology optimization methods) and mesh-based methods (such as morphing methods or mesh regulation). However, none of these methods fully satisfies the set objectives. Thus, we will introduce in this work the FEM-CsG method : Finite Element Mesh - Constructive surface Geometry. Bolstered by its strong industrial context, this method offers a response to such constraints, i.e. the possibility to represent the optimal solution by a system of CAD parameters, the possibility to adapt the FE model to the wanted analysis and the guarantee of a robust geometrical representation and mesh stability. We offer to incorporate premeshed parameterized elementary forms into a 3D sheet meshed structures. Hence, these forms are arising from a CAD parameterized elementary form library. Furthermore, the FEM-CsG method uses a set of operators acting on the mesh allowing a constant evolution of the topology guided by optimization. Therefore, even if the topology may vary, the resulting shapes comply with CAD representations by construction, a solution better reflecting the reality of optimizations performed during the preliminary development stage. The FEM-CsG method has been validated on two simple case studies in order to bring forward its reliability. Thus, with an intelligent and coherent choice of the design variables, shape optimization issues may, with a restrictive number of variables, explore an important number of shapes and topologies. Topology changes are accomplished in a continuous manner, therefore validating the FEM-CsG method to any desired analysis
Braik, William. « Détection d'évènements complexes dans les flux d'évènements massifs ». Thesis, Bordeaux, 2017. http://www.theses.fr/2017BORD0596/document.
Texte intégralPattern detection over streams of events is gaining more and more attention, especially in the field of eCommerce. Our industrial partner Cdiscount, which is one of the largest eCommerce companies in France, aims to use pattern detection for real-time customer behavior analysis. The main challenges to consider are efficiency and scalability, as the detection of customer behaviors must be achieved within a few seconds, while millions of unique customers visit the website every day,thus producing a large event stream. In this thesis, we present Auros, a system for large-scale an defficient pattern detection for eCommerce. It relies on a domain-specific language to define behavior patterns. Patterns are then compiled into deterministic finite automata, which are run on a BigData streaming platform. Our evaluation shows that our approach is efficient and scalable, and fits the requirements of Cdiscount
Laithier, Virginie. « Microsystème fluidique de détection de gaz pour l'environnement ». Thesis, Aix-Marseille, 2012. http://www.theses.fr/2012AIXM4308/document.
Texte intégralMy thesis work led to develop an innovative fluidic microsystem for gas detection. It is portable, cheap and has an integrated thermal pumping. The gas microsensor used is inserted into a microchannel. Its integrated heater allows the well detection. It also allows the creation of thermal creep phenomenon, which is at the origin of the gas flow along the channel which will be detected by the microsensor. The choices of both microchannel dimensions and the temperature gradient are the most important parameters. Thus, microfluidic and thermal simulations were performed to define the microchannel dimensions and the most suitable materials natures. Two microsystems were processed with clean room technologies. The integrated heater was calibrated to study the real thermal gradient. A study of the performance of the microsensor was performed under ammonia. We could include determining the optimal temperature sensing. Then the microsystem was studed using in a special detection cell
Kock, Guillaume. « Étude et réalisation de gyromètres à détection thermique ». Thesis, Montpellier, 2017. http://www.theses.fr/2017MONTS043/document.
Texte intégralThis PhD thesis deals with both study and fabrication of micro-sensors dedicated to the measurement of angular velocity, these devices are based on heat exchanges. Two types of gyroscopes have been developed, one using a jet of gas being deflected by the rotation, the other one using thermal expansion of a gas. Under rotation, heat transfers in a locally heated fluid is modified by the Coriolis force and induces a change in temperature distribution. Using a fluid as seismic mass enhances performances in terms shock (> 10 000g) and vibration resistances compared with standard mechanical gyroscopes.The aim of this work has been to study, develop and characterize these two gyroscopes. For this purpose, numerical studies have been carried out in order to improve our understanding of physical and thermal phenomena involved in the device. Effects of thermo-physical parameters, sensor size and gas flow velocity on both sensitivity and measuring range were analyzed.On the other hand, prototypes of both gyroscopes were manufactured and characterized. The proof of concept has been validated and encouraging results have been found. Comparison of measure and simulation have shown that improvements have to be made on the numerical model
Roudiere, Gilles. « Détection d'attaques sur les équipements d'accès à Internet ». Thesis, Toulouse, INSA, 2018. http://www.theses.fr/2018ISAT0017/document.
Texte intégralNetwork anomalies, and specifically distributed denial of services attacks, are still an important threat to the Internet stakeholders. Detecting such anomalies requires dedicated tools, not only able to perform an accurate detection but also to meet the several constraints due to an industrial operation. Such constraints include, amongst others, the ability to run autonomously or to operate on sampled traffic. Unlike supervised or signature-based approaches, unsupervised detection do not require any kind of knowledge database on the monitored traffic. Such approaches rely on an autonomous characterization of the traffic in production. They require the intervention of the network administrator a posteriori, when it detects a deviation from the usual shape of the traffic. The main problem with unsupervised detection relies on the fact that building such characterization is complex, which might require significant amounts of computing resources. This requirement might be deterrent, especially when the detection should run on network devices that already have a significant workload. As a consequence, we propose a new unsupervised detection algorithm that aims at reducing the computing power required to run the detection. Its detection focuses on distributed denial of service attacks. Its processing is based upon the creation, at a regular interval, of traffic snapshots, which helps the diagnosis of detected anomalies. We evaluate the performances of the detector over two datasets to check its ability to accurately detect anomalies and to operate, in real time, with limited computing power resources. We also evaluate its performances over sampled traffic. The results we obtained are compared with those obtained with FastNetMon and UNADA
Thierry, Aurélien. « Désassemblage et détection de logiciels malveillants auto-modifiants ». Thesis, Université de Lorraine, 2015. http://www.theses.fr/2015LORR0011/document.
Texte intégralThis dissertation explores tactics for analysis and disassembly of malwares using some obfuscation techniques such as self-modification and code overlapping. Most malwares found in the wild use self-modification in order to hide their payload from an analyst. We propose an hybrid analysis which uses an execution trace derived from a dynamic analysis. This analysis cuts the self-modifying binary into several non self-modifying parts that we can examine through a static analysis using the trace as a guide. This second analysis circumvents more protection techniques such as code overlapping in order to recover the control flow graph of the studied binary. Moreover we review a morphological malware detector which compares the control flow graph of the studied binary against those of known malwares. We provide a formalization of this graph comparison problem along with efficient algorithms that solve it and a use case in the software similarity field
Surre, Jérémy. « Détection précoce de la sensibilité bactérienne aux antibiotiques ». Thesis, Sorbonne Paris Cité, 2017. http://www.theses.fr/2017USPCB078/document.
Texte intégralFollowing the discovery of antibiotics, the therapeutic successes foreshadowed a future where infectious diseases of bacterial origin would be eradicated. However, in less than a century, the massive use of broad-spectrum antibiotics led to the emergence of resistance thus reducing therapeutic options. My research project aims to understand early bacterial metabolic and morphological changes induced by antibiotics and to contribute to the development of rapid and reliable diagnostic tests to promote the implementation of more targeted antibiotic treatments. By monitoring changes in various metabolic and morphological parameters of bacteria after antibiotic treatment, we have shown the interest of viability markers such as DiBAC4(3), TOPRO®-3 or Alexa FluorTM Hydrazide for rapid detection (<3h) of bacterial susceptibility to antibiotics. In particular, we have shown for the first time that protein carbonylation, which is induced under conditions of oxidative stress and cellular aging, is a universal early marker of bactericidal antibiotic susceptibility. Following this first part of the study, we wanted to understand the mechanisms involved in bacterial response to lethal stress caused by antibiotics. Following this first part of the study, we wanted to understand the bacterial mechanisms involved in response to lethal stress caused by antibiotics. In our experiments, it was observed that when the conditions no longer allowed the organism survival, a fluorescence signal intrinsically linked to the bacterium allowed to predict the fatal outcome after only 2 hours of incubation. Indeed, following a treatment with a bactericidal antibiotic targeting the synthesis of bacterial peptidoglycan (ampicillin), we observed a maximum fluorescence of the cells at the dose of antibiotic corresponding to the Minimum Inhibitory Concentration (MIC). The fluorescence increase of bacterial cells was also observed during the lethal treatment with a biocidal agent (sodium hypochlorite). However, this phenomenon is no longer observable with bacteriostatic or bactericidal antibiotics that inhibit protein synthesis indicating active bacterial metabolism importance. The correlations of spectral properties allowed us to suspect the flavin molecules as responsible for the observed autofluorescence phenomenon. In addition, we showed an overactivation of the biosynthesis pathway of flavin-type cofactors and flavoproteins occurring during ampicillin treatment. Finally, we performed cell sorting and cell survival experiments of ampicillin-treated bacterial populations. Our results showed that highly fluorescent cells have an average survival 5 times higher than low fluorescent cells. This suggests that the fluorescence signal observed is a cellular response mediated by flavonoid compounds in an attempt to survive to antibiotic treatment. Exploratory work suggests that the phenomenon studied in bacteria is conserved among yeasts and human cells. These results open new perspectives in bacterial physiology understanding, the study of bacterial response to exogenous stress and the rapid monitoring of cell viability
Edouard, Amosse. « Détection et analyse d’événement dans les messages courts ». Thesis, Université Côte d'Azur (ComUE), 2017. http://www.theses.fr/2017AZUR4079/document.
Texte intégralIn the latest years, the Web has shifted from a read-only medium where most users could only consume information to an interactive medium allowing every user to create, share and comment information. The downside of social media as an information source is that often the texts are short, informal and lack contextual information. On the other hand, the Web also contains structured Knowledge Bases (KBs) that could be used to enrich the user-generated content. This dissertation investigates the potential of exploiting information from the Linked Open Data KBs to detect, classify and track events on social media, in particular Twitter. More specifically, we address 3 research questions: i) How to extract and classify messages related to events? ii) How to cluster events into fine-grained categories? and 3) Given an event, to what extent user-generated contents on social medias can contribute in the creation of a timeline of sub-events? We provide methods that rely on Linked Open Data KBs to enrich the context of social media content; we show that supervised models can achieve good generalisation capabilities through semantic linking, thus mitigating overfitting; we rely on graph theory to model the relationships between NEs and the other terms in tweets in order to cluster fine-grained events. Finally, we use in-domain ontologies and local gazetteers to identify relationships between actors involved in the same event, to create a timeline of sub-events. We show that enriching the NEs in the text with information provided by LOD KBs improves the performance of both supervised and unsupervised machine learning models
Truong, Charles. « Détection de ruptures multiples – application aux signaux physiologiques ». Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLN030/document.
Texte intégralThis work addresses the problem of detecting multiple change points in (univariate or multivariate) physiological signals. Well-known examples of such signals include electrocardiogram (ECG), electroencephalogram (EEG), inertial measurements (acceleration, angular velocities, etc.). The objective of this thesis is to provide change point detection algorithms that (i) can handle long signals, (ii) can be applied on a wide range of real-world scenarios, and (iii) can incorporate the knowledge of medical experts. In particular, a greater emphasis is placed on fully automatic procedures which can be used in daily clinical practice. To that end, robust detection methods as well as supervised calibration strategies are described, and a documented open-source Python package is released.The first contribution of this thesis is a sub-optimal change point detection algorithm that can accommodate time complexity constraints while retaining most of the robustness of optimal procedures. This algorithm is sequential and alternates between the two following steps: a change point is estimated then its contribution to the signal is projected out. In the context of mean-shifts, asymptotic consistency of estimated change points is obtained. We prove that this greedy strategy can easily be extended to other types of changes, by using reproducing kernel Hilbert spaces. Thanks this novel approach, physiological signals can be handled without making assumption of the generative model of the data. Experiments on real-world signals show that those approaches are more accurate than standard sub-optimal algorithms and faster than optimal algorithms.The second contribution of this thesis consists in two supervised algorithms for automatic calibration. Both rely on labeled examples, which in our context, consist in segmented signals. The first approach learns the smoothing parameter for the penalized detection of an unknown number of changes. The second procedure learns a non-parametric transformation of the representation space, that improves detection performance. Both supervised procedures yield finely tuned detection algorithms that are able to replicate the segmentation strategy of an expert. Results show that those supervised algorithms outperform unsupervised algorithms, especially in the case of physiological signals, where the notion of change heavily depends on the physiological phenomenon of interest.All algorithmic contributions of this thesis can be found in ``ruptures'', an open-source Python library, available online. Thoroughly documented, ``ruptures'' also comes with a consistent interface for all methods
Marcadon, Frédéric. « Détection et modélisation de binaires sismiques avec Kepler ». Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLS080/document.
Texte intégralThe Kepler space telescope detected solar-like oscillations in several hundreds of stars, providing a way to determine their physical properties using asteroseismology. However, the stellar evolutionary models and scaling relations employed to determine parameters such as the mass, the radius and the age require a proper calibration. In this context, the use of seismic binaries showing solar-like oscillations in both stars is especially suitable. During this thesis, we have worked towards the detection of such seismic binaries from the Kepler database and developed the necessary tools to study them. Although the discovery of a new seismic binary was very unlikely, we were able to report for the first time the detection of solar-like oscillations in the two brightest stars of a triple stellar system, namely HD 188753. Using stellar modelling, we found compatible ages for the two stars derived from asteroseismology, as expected from their common origin. In addition, we performed the first orbital analysis of this hierarchical system in order to derive a direct estimate of masses and parallax. Finally, the example of HD 188753 shows our capability to detect and model each of the stars of a binary or multiple system and to perform the orbital analysis of this one. The various tools developed during this thesis will be extensively used in the context of the future missions TESS and PLATO
Maire, Florian. « Détection et classification de cibles multispectrales dans l'infrarouge ». Thesis, Evry, Institut national des télécommunications, 2014. http://www.theses.fr/2014TELE0007/document.
Texte intégralSurveillance systems should be able to detect potential threats far ahead in order to put forward a defence strategy. In this context, detection and recognition methods making use of multispectral infrared images should cope with low resolution signals and handle both spectral and spatial variability of the targets. We introduce in this PhD thesis a novel statistical methodology to perform aircraft detection and classification which take into account these constraints. We first propose an anomaly detection method designed for multispectral images, which combines a spectral likelihood measure and a level set study of the image Mahalanobis transform. This technique allows to identify images which feature an anomaly without any prior knowledge on the target. In a second time, these images are used as realizations of a statistical model in which the observations are described as random spectral and spatial deformation of prototype shapes. The model inference, and in particular the prototype shape estimation, is achieved through a novel unsupervised sequential learning algorithm designed for missing data models. This model allows to propose a classification algorithm based on maximum a posteriori probability Promising results in detection as well as in classification, justify the growing interest surrounding the development of multispectral imaging devices. These methods have also allowed us to identify the optimal infrared spectral band regroupments regarding the low resolution aircraft IRS detection and classification
Goix, Nicolas. « Apprentissage automatique et extrêmes pour la détection d'anomalies ». Thesis, Paris, ENST, 2016. http://www.theses.fr/2016ENST0072/document.
Texte intégralAnomaly detection is not only a useful preprocessing step for training machine learning algorithms. It is also a crucial component of many real-world applications, from various fields like finance, insurance, telecommunication, computational biology, health or environmental sciences. Anomaly detection is also more and more relevant in the modern world, as an increasing number of autonomous systems need to be monitored and diagnosed. Important research areas in anomaly detection include the design of efficient algorithms and their theoretical study but also the evaluation of such algorithms, in particular when no labeled data is available -- as in lots of industrial setups. In other words, model design and study, and model selection. In this thesis, we focus on both of these aspects. We first propose a criterion for measuring the performance of any anomaly detection algorithm. Then we focus on extreme regions, which are of particular interest in anomaly detection, to obtain lower false alarm rates. Eventually, two heuristic methods are proposed, the first one to evaluate anomaly detection algorithms in the case of high dimensional data, the other to extend the use of random forests to the one-class setting
Verdier, Loïc. « Détection de structures par combinaison des données Planck et BOSS et détection simultanée d’amas de galaxies dans les données Planck et ROSAT ». Thesis, Université Paris-Saclay (ComUE), 2016. http://www.theses.fr/2016SACLS192/document.
Texte intégralOriginating from the gravitational collapse of the primordial fluctuations of matter, galaxy clusters are the mixture of a dark matter halo, a baryonic plasma also called « hot gas » and several galaxies. Cluster counts provide stringent constraints on cosmology.Improving the detection of the hot gas component in nearby or distant structures is the main goal of my work. We can detect this hot gas in the Planck satellite maps thanks to the Sunyaev-Zel'dovich effect and in the ROSAT satellite maps by bremsstrahlung.The first part of my thesis is dedicated to the detection of hot gas in Planck maps at the positions of BOSS quasars, so in distant structures.Dust-like emission dominates our signal. I developped a new filter in order to separate the possible sub-dominant hot gas emision from the dust one. I get a significnt hot gas signal for structures with redshift z >2.5.In the second part of the thesis, the simultaneous detection, I build a filter suited for extracting clusters (z<1.5) individually in the ROSAT data respecting the noise statitistics. This filter is also designed to be combined with the filter used for detecting clusters in Planck maps. Increasing the number of less massive clusters will be the final goal
Omar, Abdelaziz. « Détection ultrasensible de molécules d'intérêts atmosphériques dans l'infrarouge lointain ». Thesis, Littoral, 2016. http://www.theses.fr/2016DUNK0430/document.
Texte intégralQThe detection of trace pollutants is an important issue for monitoring air quality. Terahertz spectroscopy, used to probe spectral regions rich in molecular absorption, is an appropriate technique for measuring atmospheric pollution. This work of thesis consisted in developing and characterizing an ultrasensitive spectrometer to measure molecules of atmospheric interest. A terahertz spectrometer was mounted using a frequency multiplication chain emitting up to 900 GHz. The detection sensitivity has been optimized and characterized. In collaboration with the IEMN, a terahertz spectrometer using a vector network analyzer transmitting up to 500 GHz as a source, has been set up and characterized. The potential of terahertz spectroscopy has been demonstrated to monitor in real time the evolution of concentrations of gaseous pollutants during a chemical reaction and to deduce the kinetic rates. Following the rotational transitions of H2CO and CO of a photolysis reaction of formaldehyde, kinetic reaction rates were determined. Detection of radicals is a challenge because of their high reactivity. The configuration of our spectrometer was adapted to optimize sensitivity and to study the photolysis reaction of acetaldehyde using "Photosensitization" by mercury. Frequency modulation and Zeeman effect modulation were used to study HCO. The sensitivity was optimized, quantified HCO and measured over 200 absorption lines. A spectroscopy study of HCO is initiated in order to optimize the parameters of the international data bases
Cosin, Charlotte. « Troubles de l'humeur post-AVC, caractérisation et détection précoce ». Thesis, Paris, EPHE, 2016. http://www.theses.fr/2016EPHE3051/document.
Texte intégralStroke is the second leading cause of death and the leading cause of acquired disability in adults worldwide. Over the last twenty years, treatment in the acute phase of stroke has improved considerably. As a result, post-stroke mortality began to decline and the proportion of surviving patients with mild or moderate disabilities has increased. The long-term monitoring of this category of patients allowed to highlight, in a significant proportion of them, the occurrence of high psychological distress. These mood disorders significantly reduce post-stroke quality of life and, therefore, the assumption that put primary emphasis on the motor deficits, sensory or language now begins to move towards the psychological care of patients. These complications are however still insufficiently understood and their management remains unsatisfactory.In this context, the objective of this PhD was to better describe these post-stroke mood disorders and highlight, through objective measurement tools, the existence of risk factors for their occurrence. 91 patients were followed during one year after stroke and were evaluated on cognitive assessments, mood state evaluation, sleep recording and language recording. The results allowed us to better understand the involvement of some variables in the occurrence or progression of some mood disorders, for example sleep fragmentation and post-stroke apathy or affective prosody and post-stroke depression. Cerebellar lesions also appear to be related to the occurrence of post-stroke mood disorders. These results are part of the increasing research on post-stroke mood disorders
Aubert, Pierre. « Calcul haute performance pour la détection de rayon Gamma ». Thesis, Université Paris-Saclay (ComUE), 2018. http://www.theses.fr/2018SACLV058/document.
Texte intégralThe new generation research experiments will introduce huge data surge to a continuously increasing data production by current experiments. This increasing data rate causes upheavals at many levels, such as data storage, analysis, diffusion and conservation.The CTA project will become the utmost observatory of gamma astronomy on the ground from 2021. It will generate hundreds Peta-Bytes of data by 2030 and will have to be stored, compressed and analyzed each year.This work address the problems of data analysis optimization using high performance computing techniques via an efficient data format generator, very low level programming to optimize the CPU pipeline and vectorization of existing algorithms, introduces a fast compression algorithm for integers and finally exposes a new analysis algorithm based on efficient pictures comparison
Jin, Lijing. « Détection des interactions photon-photon dans un circuit supraconducteur ». Thesis, Université Grenoble Alpes (ComUE), 2016. http://www.theses.fr/2016GREAY002/document.
Texte intégralSuperconducting quantum circuits are currently investigated in many labs. An important goal of this research is to build and couple several superconducting quantum bits for quantum information processing. The theoretical description of these circuits is based on the coupled dynamics of the superconducting phase differences at the Josephson junctions and the degrees of freedom of the surrounding electromagnetic environment. This nonlinear dynamics has been much studied in the classical regime. Quantum effects have been considered by perturbative methods, assuming that some degrees of freedom are thermalized. In contrast, the coupled quantum dynamics was largely unexplored in a non-perturbative regime. The goal of the trainee will be to describe the crossover between the classical and quantum regimes in the case of a single Josephson junction coupled to an electromagnetic resonator. For this, the candidate will study the master equations for this system with analytical and numerical methods. In the long term, the aim will be to understand how to generate and manipulate photons thanks to time-resolved electric control of the circuit
Nguyen, Thi Khanh Hong. « Conception faible consommation d'un système de détection de chute ». Thesis, Nice, 2015. http://www.theses.fr/2015NICE4093/document.
Texte intégralNowadays, fall detection is a major challenge in the public health care domain, especially for the elderly living alone and rehabilitants in hospitals. This thesis presents an exploration for a Fall Detection System based on camera under an algorithmic and architectural point of view. Our system includes four modules: Object Segmentation, Filter, Feature Extraction and Recognition and give an urgent alarm for detecting different kinds of fall. Firstly, different algorithms for the Fall Detection System are proposed and compared the efficiency among Background Subtraction-Neural Network, Background Subtraction-Template Matching (BGS/TM), Background Subtraction-Hidden Markov Model, and Gaussian Mixture Model. Therefore, the selected BGS/TM with 91.67% (Recall), 100% (Precision) and 95.65% (Accuracy) will be implemented on ZYNQ platform. Moreover, a DUT-HBU database which is classified with different actions: fall, non-fall in three camera directions is used to evaluate the efficiency of this system. Secondly, the aim is to explore low cost architectures for this system, new power consumption and execution time models for processor core and FPGA are defined according to the different configurations of architecture and applications. The error rates of the proposed models don’t exceed 3.5%. The models are then extended to hardware/software architectures to explore low cost architecture by defining a suitable Design Space Exploration methodology. Two techniques for parallelization which are based on intra-task and inter-task static scheduling are applied with the aim to enhance the accuracy and the power consumption of this system reaches 98.3% with energy per frame of 29.5mJ/f
Offranc, Olivier. « Composants pour la génération et la détection d’impulsions térahertz ». Thesis, Lille 1, 2010. http://www.theses.fr/2010LIL10051/document.
Texte intégralThe terahertz frequency range makes the connection between the world of transistors and lasers. Despite the many potential applications, today there is not compact terahertz source operating at room temperature. Similarly, for detection of terahertz pulses, there is not really satisfying systems for telecommunication applications. We studied and characterized low temperature growth GaAsSb (LT-GaAsSb) to obtain a resistive material with short life time for the development of photoswitches. From this material, photoconductive antennas were made for the generation of terahertz pulses. The first results are encouraging although less efficient than the LT-GaAs and could be improved by optimizing the growth of LT-GaAsSb. In addition, we made uni-travelling carrier photodiodes (UTC-PD) in GaAsSb whose response is similar to the UTC-PD in InGaAs. The monolithic integration of UTC-PD in GaAsSb with a broadband terahertz antenna enabled the generation of terahertz waves around 300GHz, thus demonstrating the feasibility of such a device for generating terahertz waves. For detection of terahertz pulses, we have designed and simulated a monostable based on resonant tunneling diode. This circuit is composed of passive elements such as an inductor to determine the width of the output pulse of the monostable. The resonant frequency of the inductor appears to be the main limitation for the detection of terahertz pulses
Brodovitch, Alexandre. « Détection et première analyse d'antigènes par les lymphocytes T ». Thesis, Aix-Marseille, 2014. http://www.theses.fr/2014AIXM4050/document.
Texte intégralT lymphocytes need to efficiently probe a large number of antigen-presenting cell (APC) in order to detect an agonist antigen and to initiate an effective immune response. Therefore, engagement of the T-cell receptor (TCR) by peptide-bound major histocompatibility complex (pMHC) is a highly sensitive and rapid process. While signaling pathways downstream the TCR are extensively studied, little is known about antigen discrimination kinetic and initial T-cell response.Model planar surfaces coated with TCR ligand allowed us to study cellular events initiated by antigen detection. Using TIRFM (total internal reflexion microscopy) and IRM (interference reflexion microscopy) we showed that T-cell initial contacts with the surface were mediated by mobile microvilli. TCR engagement triggers a pulling motion of T cell surface microvilli within seconds. Active microvilli movements are dependent on PLC-γ1 activation and on the presence of myosins (IIA, Ic and Ig) as well as cofilin and ezrin. After this extracellular environment probing period, TCR stimulation triggered a rapid and active cell spreading. Cell spreading amplitude and speed reflect the ligand activating efficacity. In less than 5 minutes pMHC with different biophysical properties were discriminated by T-cells and triggered a first specific cellular response
Bortolamiol, Tifania. « Nanotubes de carbone biparois : fonctionnalisation et détection in vitro ». Thesis, Toulouse, INPT, 2015. http://www.theses.fr/2015INPT0040/document.
Texte intégralFor many years now, carbon nanotubes (CNTs) show an important potential for biomedical applications (medical imaging, targeted drug delivery, tissue engineering, cancer treatment). All these applications involve modifications of the CNTs walls, through chemical functionalisation, in order to make possible the attachment of molecules of interest. There are two main approaches for functionalisation, involving different interactions and more or less substantial changes in CNTs. On the one hand, the most widely used is the covalent functionalisation that leads to stable covalent bonds but also modifies the structure and thus the inherent properties of CNTs. On the other hand, CNTs can be functionalised by a non-covalent way, that is, through weak interactions (hydrophobic interactions, π- π stacking, ionic interactions, etc.), which do not affect the structure of CNTs. The goals of this thesis were, in a first phase, to determine the best purification strategy for double-walled carbon nanotubes (DWNTs), synthesized by a Catalytic Chemical Vapour Deposition process developed at the CIRIMAT, in order to remove all traces of metal nanoparticles but especially of unwanted carbonaceous species (disorganised carbon); the aim was to start from the cleanest oxidised sample in which CNTs are the only form of carbon, in order to avoid any competition between different carbon species, making then impossible to get a clear conclusion. DWNTs were selected as an ideal structure which would represent both single-walled CNTs (similar morphology) and multi-walled CNTs in general. The protocols of covalent and non-covalent functionalisations of purified DWNTs were developed in a second stage. The choice of the molecules of interest to graft mainly fell on fluorophores (fluorescein isothiocyanate FITC; Cy5; a streptocyanine only fluorescent when it is covalently attached), thus allowing a direct observation of the functionalisation. Grafting ratios could be determined thanks to the combination of several characterisation techniques, which raised some questions on the stability with time of the non-covalent functionalisation, and on what is really covalently grafted on the CNTs when this way is chosen. Finally, such functionalised DWNTs were incubated with several types of cells to test their in vitro detection and the functionalisation stability in this kind of environment. Co-localisation of fluorescent DWNTs in the cells was achieved by comparison of the results obtained from confocal fluorescence and confocal Raman microscopies
Exurville, Ingrid. « Détection non destructive de modification malveillante de circuits intégrés ». Thesis, Saint-Etienne, EMSE, 2015. http://www.theses.fr/2015EMSE0800/document.
Texte intégralThe globalization of integrated circuits fabrication involves several questions about the integrity of the fabricated circuits. Malicious modifications called Hardware Trojans (HT) can be introduced during the circuit production process. Due to the complexity of an integrated circuit, it is really difficult to find this kind of alterations.This work focuses on a non-destructive method of HT detection. We use the paths delays of the studied design as a channel to detect HT. A model to describe paths delays is defined. It takes into account two important parameters which are the experimental conditions and the process variations.Faults attacks by clock glitches based on timing constraints violations have been performed to measure data paths delays. Reliable circuits are used for reference. After validating the relevance of this channel to get information on the internal behavior of the targeted design, experimental detections of HT inserted on two different abstraction levels (RTL and after place and route) were achieved. Process variations are taken into consideration in the studies to detect if the tested circuits are infected
Yakoubi, Zied. « Détection et évaluation des communautés dans les réseaux complexes ». Thesis, Paris 13, 2014. http://www.theses.fr/2014PA132045/document.
Texte intégralIn this thesis we focus, on one hand, on community detection in complex networks, and on the other hand, on the evaluation of community detection algorithms. In the first axis, we are particularly interested in Leaders driven community detection algorithms. First, we propose an enrichment of LICOD : a framework for building different leaders-driven algorithms. We instantiate different implementations of the provided hotspots. Second, we propose an extension of LICOD, we call it-LICOD. This extension introduces a self-validation step of all identified leaders. Experimental results of it-LICOD on real and artificial networks show that it outperform the initial LICOD approach. Obtained results are also competitive with those of other state-of-the art methods. Thirdly, we propose a semi-local centrality measure, called TopoCent, that address the problem of the irrelevance of local measures and high computational complexity of globalmeasures. We experimentally show that LICOD is often more efficient with TopoCent than with the other classical centrality measures. In the second axis, we propose two task-based community evaluation methods : CLE and PLE. We examine he hypothesis that the quality of community detection algorithms can be estimated by comparing obtained results in the context of other relevent tasks. The CLE approach, we use a data clustering task as a reference. The PLE method apply a link prediction task. We show that the experimentation of CLE and PLE methods gives new insights into the performance of community detection algorithms