Dissertations / Theses on the topic 'Perception auditive – Modèles mathématiques'
Create a spot-on reference in APA, MLA, Chicago, Harvard, and other styles
Consult the top 50 dissertations / theses for your research on the topic 'Perception auditive – Modèles mathématiques.'
Next to every source in the list of references, there is an 'Add to bibliography' button. Press on it, and we will generate automatically the bibliographic reference to the chosen work in the citation style you need: APA, MLA, Harvard, Chicago, Vancouver, etc.
You can also download the full text of the academic publication as pdf and read online its abstract whenever available in the metadata.
Browse dissertations / theses on a wide variety of disciplines and organise your bibliography correctly.
Khalidov, Vasil. "Modèles de mélanges conjugués pour la modélisation de la perception visuelle et auditive." Grenoble, 2010. http://www.theses.fr/2010GRENM064.
Full textIn this thesis, the modelling of audio-visual perception with a head-like device is considered. The related problems, namely audio-visual calibration, audio-visual object detection, localization and tracking are addressed. A spatio-temporal approach to the head-like device calibration is proposed based on probabilistic multimodal trajectory matching. The formalism of conjugate mixture models is introduced along with a family of efficient optimization algorithms to perform multimodal clustering. One instance of this algorithm family, namely the conjugate expectation maximization (ConjEM) algorithm is further improved to gain attractive theoretical properties. The multimodal object detection and object number estimation methods are developed, their theoretical properties are discussed. Finally, the proposed multimodal clustering method is combined with the object detection and object number estimation strategies and known tracking techniques to perform multimodal multiobject tracking. The performance is demonstrated on simulated data and the database of realistic audio-visual scenarios (CAVA database)
Caramiaux, Baptiste. "Etudes sur la relation geste-son en performance musicale." Paris 6, 2011. http://www.theses.fr/2011PA066461.
Full textBruckert, Laetitia. "Production et perception de la voix : entre données phylogénétiques et modèles socio-culturels." Paris 10, 2006. http://www.theses.fr/2006PA100077.
Full textThe present thesis focuses on voice production and perception. We used male voices. Theses are the main results: - Consensus on the voice hedonistic judgment and on the inferences regarding the speaker can be noted. Theses appear whatever the linguistic nature of the vocal production listened, even while listening to a simple series of vowels. - The listeners prove themselves able to infer correctly the speakers' age but not height, as they mistakenly use non-reliable acoustic indications such as pitch. - There is no gender effect on the judgments : both male and female listeners seem to produce the same judgments. - A vocal corpus effect on the judgments produced while listening to voices can be observed, principally opposing the vowels to the other corpora. -It generally turns out that male and female speakers use mainly voice prosodic information and not the spectral aspects such as pitch and the voice tone
Dugué, Pierre. "Vers un modèle de la chaîne auditive humaine dans le traitement de l’enveloppe temporelle." Rennes 1, 2008. http://www.theses.fr/2008REN1S024.
Full textThis work is aimed at studying the auditory perception of sound temporal envelope, i. E. The amplitude fluctuations whose frequency is lower than about fifty Hertz, this envelope being closely related to speech intelligibility. A physiological model of the temporal envelope processing is proposed. It is composed of two parts : the first one, from the outer ear to the inferior colliculus, generates unitary responses and the second one, from the medial geniculate body to the primary auditory cortex, neural populations responses. In addition to indices frequently used in the literature, two new indices are proposed and validated. Responses to amplitude modulation measured in the main processing centres of the temporal envelope, particularly in the human primary auditory cortex, are effectively reproduced in terms of modulation transfer functions
Bezerra, Silvio José. "Un modèle de perception visuelle pour la vision artificielle." Paris 6, 1993. http://www.theses.fr/1993PA066026.
Full textFarkhatdinov, Ildar. "Modeling verticality estimation during locomotion." Paris 6, 2013. http://www.theses.fr/2013PA066085.
Full textA nonlinear model of the vestibular system is proposed. The model was constructed from general considerations regarding the Newton-Euler dynamics governing the three-dimensional movements of bodies constrained to oscillate in non-inertial frames, such as the otoliths, which were modeled as spherical damped pendula. Two configurations were considered. The medial model considered only one inner ear located in the center of a head. The lateral model considered two inner ears located laterally with respect to the center of rotation of the head. An observability test was used to verify whether the reconstruction of the head orientation with respect to the gravitational vertical was possible from otoliths measurements only. It could be shown that in order for the head vertical orientation to be observable, the head had to be stabilized during locomotion. It was shown that the gravitoinertial ambiguity could be resolved if the head was horizontally stabilized. The results were applied to solve the head vertical orientation estimation problem. The simulations indicated that the estimation errors were smaller and the observers converged faster when head was stabilized during locomotion, leading to a nonlinear, combined observation-control system that could be stabilized with respect to the gravitational vertical. The results were tested with an experimental setup that comprised an actuated gimbal mechanism to represent the head-neck system and a liquid-based inclinometer that represented the otoliths organs. The findings derived from this research would be helpful for analyzing spatial perception in humans, and for improving the perceptual capabilities of robotic systems
Moffat, Graeme. "Etudes des acouphènes chez l'homme et dans des modèles animaux : corrélats électrophysiologiques chez l'animal éveillé : rôle des prothèses auditives dans le traitement des acouphènes chez l'homme." Aix-Marseille 1, 2010. http://www.theses.fr/2010AIX11007.
Full textTinnitus, the perception of sound in the absence of any stimulus, is a chronic problem for tens of millions of individuals worldwide, and no treatments have yet shown a high degree of eficacy in all cases. The development of effective treatments is therefore imperative and requires both an understanding of the physiologcal basis of tinnitus and of the effectiveness and limits of different treatment strategies. Given a well-established relationship between tinnitus and hearing impairment, and a consensus among physiological models of tinnitus as to the important role of deafferentation (reduced input) in the auditory system, treatment strategies that attempt to restore input to the central auditory system via suprathreshold acoustic stimulation should be effective in altering or alleviating this disorder. In collaboration with laboratories in Lyon and Marseille, France, I tested the effectiveness of two acoustic treatment protocols designed to alleviate tinnitus. Ln a first experiment, patients suffering from tinnitus were fitted for one of two hearing aid amplification regimes: standard clinical amplification, which is known to alleviate tinnitus in some cases, and high-bandwidth amplification, which the neurophysiological model of tinnitus predicts should be more effective than the clinical standard. L found that tinnitus sensation was affected by standard amplification but not by high-bandwidth, suggesting an unforeseen technical or physiological limit on the efficacy of this treatment strategy. Consideration must therefore be given to the influence of cochlear dead regions and sensitive periods, or lack thereof, in auditory neural plasticity. In a second series of experiments, my collaborators and I looked for neural correlates of tinnitus in an awake aninal model, which has not been done previously using multiple tinnitus induction strategies. Animals were subjected to a Zwicker Tone-inducing stimulus, a high dose of sodium salicylate, and an acute noise trauma, all of which are known to induce a phantom auditory percept over different time scales. We found no neural correlate of the Zwicker Tone phenomenon, but did note both an increase in evoked potential amplitude (in audiory cortex local field potentials) and a decrease over a broad band in the EEG power spectrum between 10-30 Hz after both salicylate and nobe trauma adminmration. Moreover, we found a generalized decrease in entropy in EEG under salicylate administration. These results suggest that an important experimental tinnitus model, salicylate administration, may not be completely analogous to the effects of noise trauma, in that important central changes independent of auditory effects may exert an influence. Taken together, these experiments suggest important gaps in the neurophysiological model of tinnitus, notably that though tinnitus induction may not necessarily depend on peripheral deafferentation (as in salicylate administration), its sustained presence is likely in most cases dependent upon peripheral or sensorineural damage at a level close to the cochlea, and that cases of peripheral (cochlear) damage may present an insurmountable limit to the effetiveness of acoustic therapies. Further implications and consideration of other treatment strategies is discussed
Leiba, Raphaël. "Conception d'un outil de diagnostic de la gêne sonore en milieu urbain." Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066602/document.
Full textNoise, especially road traffic noise, is cited by many studies as a source of major societal concern. So far, public responses are based only on energy quantification of sound exposure, often by measuring or estimating LA or Lden, and sound-level reduction related decision are taken. Nevertheless, psychoacoustic studies have shown that the sound level explains only a small part of the perceived noise annoyance. It is interesting to have more information about the source of noise and not to reduce the information to its sound level. In this thesis a tool is proposed for estimating the noise annoyance induced by each road vehicle using its audio signal and noise annoyance models. To do so, the audio signal of the vehicle is isolated by using inverse methods, large microphones arrays and image processing to obtain its trajectory. The knowledge of the trajectory and of the signal allows the vehicle to be classified by a machine learning method according to Morel et al. taxonomy. Once its category obtained, the specific annoyance of the vehicle is estimated thanks to a noise annoyance model using psychoacoustic and energetic indices. This allows the estimation of specific noise annoyance for each vehicle within the road traffic. The application of this method is made during a measurement day on a large Parisian artery
Follet, Brice. "Etude sur la dichotomie ambiant-focal du comportement oculomoteur dans la perception des scènes naturelles." Paris 8, 2012. http://octaviana.fr/document/178646946#?c=0&m=0&s=0&cv=0.
Full textThis work relies on the question of existence of two kinds of visual named ambient and focal. Document begins to present previous work at behavioral, physiological, oculometric and computational levels. Next, we expose our experimental approach and the obtained results. All experimental methodologies rely on oculomotor data with the aim to study ambient and focal aspects. We approach this problem in suggesting an automatic method to discriminate the two kinds of fixations contrary to previous studies. In a first step, this fixation clustering allows to reveal one relevant behavioral index to discriminate fixations. In a subsequent step, we observe that two kinds of fixations show distinct features on the spatial deployment in the visual scene. In a next step, we investigate the respective role of ambient and focal fixations in the perception across two different paradigms. Hypothesis suggesting focal fixations are specialized in a local perception of object and ambient fixations preferentially treat the global scene are validated. In a next step, we try to consider ambient and focal fixations to increase the ability of a saliency model to predict spatial deployment of fixations. The last step suggests a functional temporal synthesis and a global schema of ocular behavior
Gueguen, Robin. "Virtualisation architecturale visuelle et auditive du théâtre antique d'Orange." Thesis, Sorbonne université, 2018. http://www.theses.fr/2018SORUS163.
Full textThis multidisciplinary project initiated by Sorbonne University aims to provide mathematical and computing solutions to archaeological problems. It deals with the ancient theater of Orange. An extensive literature review allowed to restore this monument thanks to the software Blender. In addition to the visualization part, this project also aims to study the acoustics of the monument. Thus, a numerical simulation tool especially designed for this significant size numerical problems has been developed in C++ in partenership with the CMAP (École polytechnique). The algorithm uses geometric methods thanks to the high frequency approximation. From a sound source, beams carrying a certain amount of energy are propagated in all directions and reflected on the building walls. The method provide the reverberation curve of a room for eight octave bands by considering the materials properties and the propagation medium. The algorithm is optimized by using a Divide and Conquer approach with a hierarchical octree structure. This allow to reduce the quadratic complexity of the ray/element interactions to near-linear and significantly improves computation time. The algorithm is validated by comparison with theoretical test cases. The study of the Orange theater is eventually carried out with this acoustic simulator. Thanks to different output data, it is possible to qualify the acoustic impact of certain theater configurations : presence of ornament or roof, position of spectators or sound sources
Favali, Marta. "Formal models of visual perception based on cortical architectures." Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066094/document.
Full textThe objective of this thesis is to develop mathematical models of visual perception based on cortical architectures and to apply them to reproduce phenomenological experiments and to process natural images. We primarly focus on low level vision tasks and in particular we are interested in the problem of grouping and of individuation of perceptual units. In this setting we will face the problem of the reconstruction of illusory figures and the detection of retinal vessels in optical images. Then we consider the problem of encoding and decoding of the fMRI signal from in vivo acquired brain activity of visual cortex. This allows to estimate the structure of the cortex of a specific human patient and eventually to reconstruct the visual stimulus from fMRI activity, in a so called “brain reading” strategy. The difference between our approach and the state of the art literature consists in using previously defined neuromathematical models of the cortices as a-priori knowledge to regularise in vivo estimated structure. Even if it is a long term objective, we propose a first approach to improve the results in this field. The entire work of this thesis has been developed taking into account results from phenomenology of perception and results of neurophysiology.In the field of the phenomenology of perception, at the beginning of the last century, the theory of the Gestalt psychology [Wertheimer, 1938, Kohler, 1947, Kofka, 1935] defined the integration of contours and in particular they defined grouping laws underlying perception. These are crucial in the construction of visual objects: points with characteristics in common can be grouped together to form a new visual object. Many psychophysical experiments have been proposed to measure the quantitative parameters of these laws. A particular interest of this thesis is the concept of association fields introduced by Field et al. [1993] which encodes different Gestalt principles (as good continuation, proximity). They showed that stimulus co-linearity and co-circularity play an important role for the feature of grouping. Their study showed how chances of perceiving the curvilinear path were high if the orientation of its features was the one tangent at that point and collapsed as their relative orientation deviated from being tangent
Jouve, Bertrand. "La multiplicité des aires visuelles du cortex : approche par la théorie des graphes." Paris, EHESS, 1999. http://www.theses.fr/1999EHESA007.
Full textLe, Gallou Sylvain. "Détection robuste des éléments faciaux par modèles actifs d'apparence." Rennes 1, 2007. http://www.theses.fr/2007REN1S083.
Full textFor man-machine interfaces, interactions with the machines in an unconstrained environment are a major issue. We use the Active Appearance Models (AAM) to precisely localize the eyes, the nose and the mouth of faces. Our work consists of making the AAM more robust to illumination, pose, identity and expression of faces. On the one hand, we propose a pretreatment based on oriented maps to get independent from the effects of illumination variations and on the other hand, an adaptive system allowing the AAM to focus itself in real time on the pre-learned specific model of a face the more adapted to the analyzed unknown face
Quintard, Ludovic. "Evaluation de la qualité des dispositifs d'affichage couleur : des évaluations subjectives à la mesure objective." Poitiers, 2008. http://theses.edel.univ-poitiers.fr/theses/2008/Quintard-Ludovic/2008-Quintard-Ludovic-These.pdf.
Full textThis thesis was conducted in cooperation with the Laboratoire National de Métrologie et d’Essais (LNE) within the framework of a CIFRE convention. This work consisted in realising an objective metric dedicated to the assessment of image retrieval devices (screens). To address the underlying issues, we developed an algorithm for estimating quality using the monitor features and those of the Human Visual System. The developed algorithm is made up of three blocks. The first block takes into account the technical features of the monitor (transfer function, gamut, etc. ) to turn a digital image (sent to the monitor using a digital device) into a physical image (displayed by the monitor). The second block of the algorithm models the transformation of the physical image into a perceptual image. This modelling involves familiarity with human visual system operational principles, particularly the encoding of colour information, as well as the perceptual attributes involved in judging a colour digital image quality. The required knowledge is described in chapter 4 of the present thesis. The third and last block was developed to rate the quality of this perceptual image. A subjective study of the quality of 5 different LCD screens involving thirty participants and based on six analysis criteria was performed. It successfully showed that colour, saturation and contrast are major factors on which the screen quality is assessed. The metric developed was based on these findings and shows good correlation with subjective evaluation data, thus showing similarity with human judgement. The metric was also tested in the field of JPEG and JPEG2000 compression with similar efficiency
Al, Mohammad Hussam A. "Contribution à la coordination commande-perception tactile dans les systèmes d'assemblage automatisés." Paris 11, 1985. http://www.theses.fr/1985PA112086.
Full textIn this thesis, we describe the dynamic command of a manipulator having six degrees of freedom. This command is shown to have kinematic and dynamic optimal properties. In part 2 a trajectory generation algorithm is presented. This algorithm is used for a manipulator or robot for non-null speed movements in Cartesian space. In the third and the fourth parts, we describe the tactile sensor which has been developed, and its use in grasping and in identifying object images
Boutkhil, Latifa. "Coopération entre les aires corticales pour l'acquisition des capacités de reconnaissance visuelle invariante : modélisation fonctionnelle." Paris, EHESS, 2002. http://www.theses.fr/2002EHES0121.
Full textWe can recognize objects despite changes of point of view, eye’s position, size, orientation, relative position or non rigid transformations of the object itself (for instance of a newspaper or a gymnast). How this cognitive ability can be learned? That’s the question we try to answer in that work, which slots in the Cognitive Science framework, coupling neuroscience, experimental psychology, and connectionist modelling in order to take into account the richness of the biological neural substrate and of the multiplicity of the infant’s learning. We try to describe a progressive use of different kinds of sensorial and motor information, from the maturation logic of the nervous system, in the same connectionist neural network, which’s combinatory resembles in the most closer way possible to the actual visual cortex system. Within the framework of this connectionist neural network, we focus on the problem of the acquisition of perceptive visual invariants, that we modularise in a series of different learning stages from the developmental data, and we are interested in the causal sequence generated by this network, linking for instance the development of ocular exploration and the development of infant’s perceptual abilities. The first chapter reminds the conceptual foundations of connectionism, pointing particularly on the relative invariance capacities and the limits of different “classical” neural networks models. In the first part of the second chapter, we propose a review of the data from neurobiology, experimental psychology relative to the architecture of the cortical visual system, analysed from the point of view of the objects coding for an invariant recognition. A foreword to this part will present a synthesis on the principal invariant recognition theories. The second part gives a review of the connectionist solutions to the invariant visual recognition problem, naming a classification of different kinds of neural networks models, biologically plausible or not, arranged in four big classes to get perceptive invariance: I) invariance through the input coding (local or global transformation), II) invariance thanks to changes of the neural network structure or correlation methods: III) invariance to perspective by interpolation between a collection of 2D views: a)memorization of prototypes by RBF connectionists networks and b) use of the information of the spatiotemporal continuity. With the concepts of the first, the third chapter focus on the neural processing, realized by the visual cortex, considered as an architecture of a network of “cortical column” networks. Within this connectionist paradigm, a functional benchmarking of the invariance capacities of such a model of the cortical visual system is proposed from two simulations on a network of transputers applied a task of characters recognition. The relative translation and scale invariance capacities obtained, result principally from the cooperation between two networks (the first one models the temporal way of the visual cerebral cortex, dedicated more particularly to the identification task, the second one models the parietal, dedicated to the perception of space and ocular exploration). In order to reach the goal of this thesis, which consists in finding the correspondences between the development stages of the infant’s visual system and the setting of functional relations that could allow perceptive invariance, the last chapter proposes a functional modelling, that posits at the different levels of the architecture of the visual system in maturation, the neural networks models detailed previously in the second part of the second chapter to solve the problem of invariant recognition. This functional modelling make reference to the mechanisms simulated in the second part of the third chapter
Deleforge, Antoine. "Acoustic Space Mapping : A Machine Learning Approach to Sound Source Separation and Localization." Thesis, Grenoble, 2013. http://www.theses.fr/2013GRENM033/document.
Full textIn this thesis, we address the long-studied problem of binaural (two microphones) sound source separation and localization through supervised leaning. To achieve this, we develop a new paradigm referred as acoustic space mapping, at the crossroads of binaural perception, robot hearing, audio signal processing and machine learning. The proposed approach consists in learning a link between auditory cues perceived by the system and the emitting sound source position in another modality of the system, such as the visual space or the motor space. We propose new experimental protocols to automatically gather large training sets that associates such data. Obtained datasets are then used to reveal some fundamental intrinsic properties of acoustic spaces and lead to the development of a general family of probabilistic models for locally-linear high- to low-dimensional space mapping. We show that these models unify several existing regression and dimensionality reduction techniques, while encompassing a large number of new models that generalize previous ones. The properties and inference of these models are thoroughly detailed, and the prominent advantage of proposed methods with respect to state-of-the-art techniques is established on different space mapping applications, beyond the scope of auditory scene analysis. We then show how the proposed methods can be probabilistically extended to tackle the long-known cocktail party problem, i.e., accurately localizing one or several sound sources emitting at the same time in a real-word environment, and separate the mixed signals. We show that resulting techniques perform these tasks with an unequaled accuracy. This demonstrates the important role of learning and puts forwards the acoustic space mapping paradigm as a promising tool for robustly addressing the most challenging problems in computational binaural audition
Tlapale, Olivier Émilien. "Modelling the dynamics of contextual motion integration in the primate." Nice, 2011. https://tel.archives-ouvertes.fr/tel-00850265.
Full textThis thesis addresses the study of motion integration in the primate. Based on anatomical and functional knowledge of two cortical areas involved in motion perception, namely VI and MT, we explain various perceptual and oculo-motor responses found in the literature. First, we build a recurrent model of motion integration where a minimal number of cortical interactions are assumed. Proposing a simple readout mechanism, we are able to reproduce not only motion perception but also the dynamics of smooth pursuit eye movements on various line figures and gratings viewed through different apertures. Second, following perceptual studies concerning motion integration and physiological studies of receptive fields, we construct another dynamical model where motion information is gated by form cues. To this end, we postulate that the visual cortex takes advantage of luminance smoothness in order to gate motion diffusion. Such an elementary diffusion mechanism allows to solve various contextual problems where extrinsic junctions should be eliminated, without relying on complex junction detectors or depth computation. Finally, we rewrite the initial dynamical model into the neural field formalism in order to mathematically analyse its properties. We incorporate the multiplicative feedback term into the formalism, and prove the existence and uniqueness of the solution. To generalise the comparison against visual performance, we propose a new evaluation methodology based on human visual performance and design a database of image sequences taken from biology and psychophysics literature. Offering proper evaluation methodology is essential to continue progress in modelling the neural mechanisms involved in motion processing. To conclude, we investigate the performances of our neural field model by comparison against state of the art computer vision approaches and sequences. We find that, despite its original objective, this model gives results comparable to recent computer vision approaches of motion estimation
Souissi, Nissrine. "Apport dynamique et transparent de sémantique lors de manipulations de documents." Paris 12, 2006. https://athena.u-pec.fr/primo-explore/search?query=any,exact,990003939380204611&vid=upec.
Full textThis thesis studies the conceptual modelling and the handling of data in the decisional spatial systems. Our model proposes the concept of semantic dependencies (SD) which extends the concept of functional dependencies (FD) in the data bases. The FD handles only part of the semantics of information system data, contrary to SD which make it possible to model the various semantic bonds being able to exist between the alphanumeric attributes and spatial representation. Our model brings a semantic richer than the couple (attribute, domain). The definition of an attribute becomes a triplet (attribute, domain, semantic constants). Handling relates to the retrieval of the relevant attributes and their integration in the scheme. This process of enrichment is based on the SD concept. The procedure presented by this concept has for objective to automate the operation of determination of the relevant attributes for a request
Carnec, Mathieu. "Critères de qualité d'images couleur avec référence réduite perceptuelle générique." Nantes, 2004. http://www.theses.fr/2004NANT2044.
Full textDigital images are widely used in information media. For images that target human observers, it is necessary to hold methods that can assess image quality in order to adapt image processing (image compression, image enhancement) and measure the quality of transmission services. To be efficient, these methods (called "quality criteria") must produce quality scores in close relation with subjective quality scores given by human observers (during subjective quality evaluation tests). We propose new quality criteria based on a functional and organisational model of the human visual system. This model describes the different stages of vision : from the eye to the visual cortex which contains the V1 and V2 areas, the ventral pathway and the dorsal pathway. The novelty of these new quality criteria resides in the extraction, from an image represented in a perceptual space, of features that can be compared to the ones used by the human visual system. A similarity measure between features from an original image and features from a test image (whom quality is to be assessed) enables to produce the quality score. Several similarity measures have been tested, each one using different features and so, leading to the development of a particular quality criterium. These features constitute a reduced reference of the image which can, in a transmission context, be transmitted with the distorted image so that the quality of this latter can be assessed. Results show, on two different scored images bases (containing both images and their subjective quality scores), an important correlation between some similarity measures that have been tested and subjective scores. The size of reduced references is flexible and a quantification of this reduced reference lowers its size while preserving good performances. Our work has been integrated in two applications in order to assess images quality and to compress images while choosing the quality of output images
Escobar-Zuniga, María-José. "Modèles bio-inspirés pour l'estimation et l'analyse de mouvement : reconnaissance d'actions et intégration du mouvement." Nice, 2009. http://www.theses.fr/2009NICE4050.
Full textThis thesis addresses the study of the motion perception in mammals and how bio-inspired systems can be applied to real applications. The first part of this thesis relates how the visual information is processed in the mammal's brains and how motion estimation is usually modeled. Based on this analysis of the state of the art, we propose a feedforward V1-MT core architecture. This feedforward V1-MT core architecture will be a basis to study two different kinds of applications. The first application is human action recognition, which is still a challenging problem in the computer vision community. We show how our bio-inspired method can be successfully applied to this real application. Interestingly, we show how several computational properties inspired from motion processing in mammals, allow us to reach high quality results, which will be compared to latest reference results. The second application of the bio-inspired architecture proposed in this thesis, is to consider the problem of motion integration for the solution of the aperture problem. We investigate the role of delayed V1 surround suppression, and how the 2D information extracted through this mechanism can be integrated to propose a solution for the aperture problem. Finally, we highlight a variety of important issues in the determination of motion estimation and additionally we present many potential avenues for future research efforts
Bonzom, Pierre. "Etude, simulation et réalisation de capteurs d'inspiration animale." Montpellier 2, 1998. http://www.theses.fr/1998MON20105.
Full textFranceschiello, Benedetta. "Cortical based mathematical models of geometric optical illusions." Thesis, Paris 6, 2017. http://www.theses.fr/2017PA066131/document.
Full textThis thesis presents mathematical models for visual perception and deals with such phenomena in which there is a visible gap between what is represented and what we perceive. A phenomenon which drew the interest most is amodal completion, consisting in perceiving a completion of a partially occluded object, in contrast with the modal completion, where we perceive an object even though its boundaries are not present [Gestalt theory, 99]. Such boundaries reconstructed by our visual system are called illusory contours, and their neural processing is performed by the primary visual cortices (V1/V2), [93]. Geometric models of the functional architecture of primary visual areas date back to Hoffman [86]. In [139] Petitot proposed a model of single boundaries completion through constraint minimization, neural counterpart of the model of Mumford [125]. In this setting Citti and Sarti introduced a cortical based model [28], which justifies the illusions at a neural level and provides a neurogeometrical model for V1. Another class of phenomena are Geometric optical illusions (GOIs), discovered in the XIX century [83, 190], arising in presence of a mismatch of geometrical properties between an item in object space and its associated percept. The fundamental idea developed here is these phenomena arise due to a polarization of the connectivity of V1/V2, responsible for the misperception. Starting from [28] in which the connectivity building contours in V1 is modeled as a sub-Riemannian metric, we extend it claiming that in GOIs the cortical response to the stimulus modulates the connectivity of the cortex, becoming a coefficient for the metric. GOIs will be tested through this model
Flandin, Grégory. "Modélisation probabiliste et exploration visuelle autonome pour la reconstruction de scènes inconnues." Phd thesis, Université Rennes 1, 2001. http://tel.archives-ouvertes.fr/tel-00843884.
Full textShi, Xiaobing. "Etude et mise en œuvre d'une approche unifiée pour la modélisation volumique-surfacique." Compiègne, 1991. http://www.theses.fr/1991COMPD394.
Full textWe have studied and worked out a new uniform approach for the volume and surface geometric modelling of solids. The volumes and surfaces informations are integrated in a unified model of solid for representing the shapes geometric of solid objects. The acquisition, the modelisation, the representation and the handing of surfaces informations will be the general approach in this thesis. It allows the integration of surfaces informations in a simple way, avoiding duplication in the acquisition and recording phases. Our study was in principle focused on the introduction, the representation and manipulation of surfaces informations in order to take account of the precision and the internal representation of informations associated to their tolerance. The efficiency is the essential way to bind friendly use of the modelling system
Balasque, Jean-Michel. "Contribution à l'analyse de l'influence de l'image de l'entreprise fournisseur sur la formation des préférences en milieu industriel." Bordeaux 4, 2005. http://www.theses.fr/2005BOR40017.
Full textBlusseau, Samy. "On salience and non-accidentalness : comparing human vision to a contrario algorithms." Thesis, Cachan, Ecole normale supérieure, 2015. http://www.theses.fr/2015DENS0042/document.
Full textThe present dissertation compares the human visual perception to computer vision algorithms based on a mathematical model called a-contrario theory. To this aim, it focuses on two visual tasks that are at the same time easy to model and convenient to test in psychophysical experiments. Both tasks consist in the perceptual grouping of oriented elements, namely Gabor patches. The first one is the detection of alignments and the second one extends to curves, that is to say to more general arrangements of elements in good continuation. In both cases, alignments and curves, psychophysical experiments were set up to collect data on the human visual perception in a masking context.The non-accidentalness principle states that spatial relations are perceptually relevant when their accidental occurrence is unlikely. The a-contrario theory is a formalization of this principle, and is used in computer vision to set detection thresholds accordingly. In this thesis, the a-contrario framework is applied in two practical algorithms designed to detect non-accidental alignments and curves respectively. These algorithms play the part of artificial subjects for our experiments.The experimental data of human subjects is then compared to the detection algorithms on the very same tasks, yielding two main results. First, this procedure shows that the Number of False Alarms (NFA), which is the scalar measure of non-accidentalness in the a-contrario theory, strongly correlates with the detection rates achieved by human subjects on a large variety of stimuli. Secondly,the algorithms' responses match very well the average behavior of human observers.The contribution of this thesis is therefore two-sided. On the one hand, it provides a rigorous validation of the a-contrario theory's relevance to estimate visual thresholds and implement visual tasks in computer vision. On the other hand, it reinforces the importance of the non-accidentalness principle in human vision.Aiming at reproducible research, all the methods are submitted to IPOL journal, including detailed descriptions of the algorithms, commented reference source codes, and online demonstrations for each one
Bonnasse-Gahot, Laurent. "Modélisation du codage neuronal de catégories et étude des conséquences perceptives." Paris, EHESS, 2009. http://www.theses.fr/2009EHES0102.
Full textAt the crossroads between theoretical neuroscience and psycholinguistics, this dissertation deals with the neural coding of categories and aims at studying the perceptual consequences resulting from an optimized representation. The focus is on situations where categorization is difficult due to overlapping of categories in stimulus space (a system of vowels, for example). Taking advantage of a neurobiological interpretation of the so-called 'exemplar models' originally introduced in psychology, and using mathematical tools from information theory, this work proposes an analytic study of the coding efficiency of a neuronal population with respect to a discrete set of categories. Introducing a perceptual distance based on the Kullback-Leibler divergence between, patterns of neural activity evoked by two different stimuli, it is shown not only that categorical perception naturally emerges from category learning, but also that several prototypical effects (the magnet effect, for instance) result from an optimized representation. A plausible model of information decoding is finally proposed, and reaction times during an identification task are characterized analytically. The obtained formula gives a «relationship between discrimination accuracy and response time. In the present work, all the analytical results, that are derived are supported by numerical studies as well as by qualitative and quantitative comparisons with experimental data available in both the neuroscience and the psycholinguistics Iiterature
Girard, Joffrey. "Modélisation de l'éblouissement d'inconfort dans des situations dynamiques." Thesis, Paris Est, 2019. http://www.theses.fr/2019PESC2058.
Full textThe discomfort glare is defined as the sensation of discomfort felt by an observer in presence of light sources, without necessary impacting his visual performance. This PhD investigates this phenomen on in the context of outdoor lighting. Although many models of discomfort glare have been proposed in the literature, there is no consensus in outdoor lighting. Moreover, most of these models have been developed to predict the discomfort glare level from one static light source in the visual field. In outdoor lighting, the road user is usually moving with multiple sources around him. To study this phenomenon in dynamic multi-source situations, seven psychovisual experimentations were carried out in the laboratory with panels of participants. These experimentations proposed multi-source stimuli having their characteristics (luminance, apparent size, position) that periodically varied. From the experimental data, a model was developed which predict the probability of four discomfort glare levels. It depends on the photometric and geometric characteristics of the visual scene
Marino, Caterina. "The perception of relevant surface cues shapes language development : evidence from typical and atypical populations." Thesis, Sorbonne Paris Cité, 2018. http://www.theses.fr/2018USPCB174/document.
Full textThe purpose of this work is to explore how infants perceive surface features of language at different levels of processing. Specifically, in both typical and atypical populations, we examined 1) the processing of low-level auditory cues and its relationship with later language outcomes and 2) the sensitivity to word frequency to create abstract linguistic representations of lexical categories and their relative word order. Accordingly to bootstrapping models learners are able to extract abstract, structural and hence directly unobservable properties of the target language from perceptually available surface cues in the input that correlate with the underlying structure. Indeed, infants are sensitive to certain acoustic and phonological properties of the speech input, which in turn correlate with specific grammatical/syntactic structures. In order to map the link between the perception of these low-level, surface cues and more abstract grammatical knowledge, this work is organized in two main parts. In the first part, two longitudinal studies are reported. Each infant received an auditory discrimination threshold task (using the rapid auditory processing paradigm) followed by a habituation/visual novelty detection task used as a control for general cognitive skills. The auditory discrimination threshold was evaluated using non-linguistic (tones) sounds in one cohort of infants and linguistic sounds (syllables) in another cohort of infants at 9 months in order to investigate the language-specificity of the process within the auditory modality. Subsequently, infants' vocabulary was assessed at 12-14-18 and 24 months and a cognitive test (Mullen scale) was performed at 18-20 months as another control measure for early processing competence. Results show that early processing abilities are predictive of later vocabulary size in typical infants. Importantly, atypical participants exhibited slower and less efficient processing abilities in both visual and acoustic modality. In the second part, the role of word frequency in bootstrapping the basic lexical categories of function and content words and their relative order is explored. The two lexical categories differ in their linguistic functions, phonological makeup and frequency of occurrence. Thus, their frequency-based discrimination could constitute a powerful initial mechanism for infants to acquire the basic building blocks of language. As functors constitute closed classes, while content words come in open classes, we examined whether 8 month-old French monolinguals relied on word frequency to categorize and track functors as non-replaceable items in a closed class, and content words as freely replaceable items in open classes. In five artificial grammar-learning experiments we have found that infants treat frequent words as belonging to closed classes, and infrequent words as belonging to open classes and they map the relative order of these categories onto the basic word order of their native language, French, a functor-initial language. Importantly, atypical participants showed lower ability of discrimination, encoding and memory when compared to typically developing peers. Overall this work contributes to a better understanding of the perceptual abilities that directly contribute to language development. Moreover, it proposes possible behavioural markers that can be potentially useful in the early identification of atypical learners
Yu, Chunlei. "Contribution to evidential models for perception grids : application to intelligent vehicle navigation." Thesis, Compiègne, 2016. http://www.theses.fr/2016COMP2293.
Full textFor intelligent vehicle applications, a perception system is a key component to characterize in real-time a model of the driving environment at the surrounding of the vehicle. When modeling the environment, obstacle information is the first feature that has to be managed since collisions can be fatal for the other road users or for the passengers on-board the considered vehicle. Characterization of occupation space is therefore crucial but not sufficient for autonomous vehicles since the control system needs to find the navigable space for safe trajectory planning. Indeed, in order to run on public roads with other users, the vehicle needs to follow the traffic rules which are, for instance, described by markings painted on the carriageway. In this work, we focus on an ego-centered grid-based approach to model the environment. The objective is to include in a unified world model obstacle information with semantic road rules. To model obstacle information, occupancy is handled by interpreting the information of different sensors into the values of the cells. To model the semantic of the navigable space, we propose to introduce the notion of lane grids which consist in integrating semantic lane information into the cells of the grid. The combination of these two levels of information gives a refined environment model. When interpreting sensor data into obstacle information, uncertainty inevitably arises from ignorance and errors. Ignorance is due to the perception of new areas and errors come from noisy measurements and imprecise pose estimation. In this research, the belief function theory is adopted to deal with uncertainties and we propose evidential models for different kind of sensors like lidars and cameras. Lane grids contain semantic lane information coming from lane marking information for instance. To this end, we propose to use a prior map which contains detailed road information including road orientation and lane markings. This information is extracted from the map by using a pose estimate provided by a localization system. In the proposed model, we integrate lane information into the grids by taking into account the uncertainty of the estimated pose. The proposed algorithms have been implemented and tested on real data acquired on public roads. We have developed algorithms in Matlab and C++ using the PACPUS software framework developed at the laboratory
Lezama, José. "On grouping theory in dot patterns, with applications to perception theory and 3D inverse geometry." Thesis, Cachan, Ecole normale supérieure, 2015. http://www.theses.fr/2015DENS0009/document.
Full textThis thesis studies two mathematical models for an elementary visual task: theperceptual grouping of dot patterns. The first model handles the detection ofperceptually relevant arrangements of collinear dots. The second model extendsthis framework to the more general case of good continuation of dots. In bothcases, the proposed models are scale invariant and unsupervised. They aredesigned to be robust to noise, up to the point where the structures to detectbecome mathematically indistinguishable from noise. The experiments presentedshow a good match of our detection theory with the unmasking processes takingplace in human perception, supporting their perceptual plausibility.The proposed models are based on the a contrario framework, a formalization ofthe non-accidentalness principle in perception theory. This thesis makes twocontributions to the a contrario methodology. One is the introduction ofadaptive detection thresholds that are conditional to the structure's localsurroundings. The second is a new refined strategy for resolving the redundancyof multiple meaningful detections. Finally, the usefulness of the collinear point detector as a general patternanalysis tool is demonstrated by its application to a classic problem incomputer vision: the detection of vanishing points. The proposed dot alignmentdetector, used in conjunction with standard tools, produces improved resultsover the state-of-the-art methods in the literature.Aiming at reproducible research, all methods are submitted to the IPOL journal,including detailed descriptions of the algorithms, commented reference sourcecodes, and online demonstrations for each one
Tancogne-Dejean, Manuela. "Anticiper la crise : ingénierie d'évacuation des bâtiments et management de l'évacuation, principes et éléments." Troyes, 2012. http://www.theses.fr/2012TROY0038.
Full textEnsuring people’s safety in case of fire requires evacuating the building. In the current societal context, evacuation must be rethought. The objective is to propose two tools with which to anticipate crisis that could occur in people facing an emergency based on human behaviour and the evacuation process. The Accessibility law of 2005 highlighted that we will all be affected by it at one time or another. The presence of more or less toxic fumes, impeding obstacles, and confusing noises. . . Have a negative impact on our abilities and faculties and complicate the evacuation process. The approach is based on a study of the perception of risk by elderly persons and of persons with disabilities, so as to be able to transpose the experience to all, which, as it is agreed that the phase of awareness of danger is the most important in the evacuation process. Three different methods were used: a qualitative method of questionnaire analysis to understand the perception of risk, a systemic approach to understand the process of evacuation in its entirety and a cindynic approach to determine the sources of danger. The results lead to the characterisation of the structuring factors and bias engaged in risk perception, to creating a static and dynamic modelling of the evacuation process and to propose two crisis anticipation tools: Building evacuation engineering and Evacuation management
Chéron, Jean-Baptiste. "Modélisation moléculaire de la perception de la saveur sucrée : approches structurales et dynamiques." Thesis, Université Côte d'Azur (ComUE), 2017. http://www.theses.fr/2017AZUR4066/document.
Full textSugar overconsumption is a risk factor for pathologies such as type II diabetes or obesity. Sweeteners consumption is used to overcome this public health issue. Indeed, they have low caloric value but still preserve the pleasure of sweet taste. Currently, number of sweeteners are commercially available, but they present a bitter aftertaste or there is a debate about their safety. One aim of this work was to propose new intense sweeteners using computational modeling strategies. Through a statistical approach to predict the sweetness based on the chemical structure of already known sweeteners, new natural compounds have been identified. Furthermore, the structural study of the homology model of the sweet taste receptor provides some clues to design new sweeteners. The molecular dynamic study of a class C G-protein coupled receptor gives the first molecular hypothesis of the activation process
Saval, Arnaud. "Modèle temporel, spatial et sémantique pour la découverte de relations entre événements." Caen, 2011. http://www.theses.fr/2011CAEN2098.
Full textThe popularity of social networks and new forms of communication has led to the emergence of new sources of information that should be studied. Anyone is able to publish and highlight information of interest. Today, these behaviors appear as ways to track a topic of interest: eg H1N1. However, the automatic processing of such information needs to be improved in order to define semantically a topic of interest (the implications of the Tsunami in Myanmar). This thesis propose a semantic extension of the modeling of events in time and space to represent the evolution of these topics of interest. This report explains how to use the introduced formalism to define the methods of reasoning on a knowledge base structured to improve the representation of the situation by discovering relationships in this information
Hamlaoui, Soumya. "Suivi 2D de visages et de mouvements faciaux." Compiègne, 2006. http://www.theses.fr/2006COMP1531.
Full textTracking a face and its facial features in a video sequence is a challenging problem in computer vision. This thesis addresses the problem of tracking the 2D global motion of a face as well as the local motion of its inner features. Ln this thesis, we propose a stochastic approach based on a particle filtering scheme. The observation distribution is derived from an active appearance model, or from an on-Iine estimated appearance model. The dynamics describing the state system evolution are guided by a deterministic research. The particle number is adjusted to the effective needs of the tracking at each time step. This adjustment allows an optimization of the computing time. We also use a robust distance measure which increases the tracking robustness when the face is partially occulted. Experiments on real video show encouraging results
Edjekouane, Idir. "La sonie des signaux non-stationnaires et la sonie binaurale : application à la téléphonométrie." Thesis, Aix-Marseille, 2016. http://www.theses.fr/2016AIXM4722/document.
Full textDuring a telephone conversation between two people, the perceived sound level (i.e. loudness) must be comfortable for the listening. For over thirty years, we have been using a measure called « Loudness Rating (LR) » to adjust the loudness in this comfort zone. This indicator was developed for the landline telephony. However, with recent advances in telecommunications, the current use of LR exhibits some limitations. Thus, the aim of this PhD thesis is to study the possibility of replacing the LR indicator by a more sophisticated state-of-art loudness model based on Zwicker’s model. We selected and evaluated the behavior of the four most common loudness models. For that purpose, we designed a new method of loudness assessment called « Mesure Indirecte du Niveau d’Isosonie (MINI) ». The conclusion of this evaluation is that these loudness models can be used to replace the LR in hands-free mode. However, their use in handset mode requires additional studies to improve their predictions
Muller, Jean-Denis. "La perception structurante : apprentissage non monotone de fonctions visuelles par croissance et maturation de structures neuromimétiques." Toulouse, ENSAE, 1993. http://www.theses.fr/1993ESAE0030.
Full textDrira, Hassen. "Statistical computing on manifolds for 3D face analysis and recognition." Thesis, Lille 1, 2011. http://www.theses.fr/2011LIL10075/document.
Full textAutomatic face recognition has many benefits over other biometric technologies due to the natural, non-intrusive, and high throughput nature of face data acquisition. Thus, the techniques for face recognition have received a growing attention within the computer vision community over the past three decades. In terms of a modality for face imaging, a major advantage of 3D scans over 2D color imaging is that variations in illumination and scaling have less influence on the 3D scans.However, scan data often suffer from the problem of missing parts dueto self-occlusions or imperfections in scanning technologies. Additionally, variations in face data due to facial expressions are challenging to 3D face recognition. In order to be useful in real-world applications, 3D face recognition approaches should be able to successfully recognize face scans even in the presence of large expression-based deformations and missing data due to occlusions and pose variation. Most recent research has been directed towards expression-invariant techniques and spent less effort to handle the missing parts problem. Few approaches handles the missing part problem but none has performed on a full database containing real missing data, they simulate some missing parts. We present a common framework handling both large expressions and missing parts due to large pose variation. In addition, with the same framework, we are able to average surfaces and hierarchically organize databases to allow efficient searches. In presence of occlusion, we propose to delete and restore occluded parts. The surface is first represented by radial curves (emanating from the nose tip fo the 3D face). Then a base is built using PCA for each curve. Hence, the missing part of the curve can be restored by projecting the existing part of it on the base. PCA is applied on the tangent space of the mean curve as it is linear space. Once the occlusion was detected and removed, the occlusion challenge can be handled as a missing data problem. Hence, we apply the restoration framework and then apply our radial-curve-based 3D face recognition algorithm
Vivacqua, Adriana Santarosa. "Une approche pour la dissémination d'informations dans environnements coopératifs." Compiègne, 2007. http://www.theses.fr/2007COMP1690.
Full textThis thesis presents a model and a method for information distribution in loosely structured collaborative environments. These environments are characterized by their flexibility and easy with which participants alter rules and procedures, adjusting to external events. However, this reorganization is only possible if participants have information leading to adequate decision making and ajustements between group members. Based on sociological theories, we present an agent based peer-to-peer architecture for information distribution, and a method to analyse interactions between users and discover ongoing collaboration. An analysis of this method shows it is a promising avenue, even though improvements are necessary for it function perfectly
Moras, Julien. "Grilles de perception évidentielles pour la navigation robotique en milieu urbain." Phd thesis, Université de Technologie de Compiègne, 2013. http://tel.archives-ouvertes.fr/tel-00866300.
Full textChevriaux, Yann. "Une approche qualitative spatiale pour une description sémantique des reliefs." Phd thesis, Université Rennes 1, 2008. http://tel.archives-ouvertes.fr/tel-00315651.
Full textNous cherchons à décrire une silhouette - i.e., une coupe de terrain ou la séparation terre/ciel à l'horizon - selon la perception que peut en avoir un observateur. Nous introduisons un modèle, fondé sur une approche qualitative, qui consiste à décrire une silhouette par une séquence de symboles signifiants. L'utilisateur ayant la possibilité de définir ses propres catégories, le modèle possède la capacité de s'adapter à différents contextes.
L'originalité de notre modèle repose dans la méthode de détection des formes significatives. Nous nous ecartons volontairement des méthodes numériques généralement utilisées dans les systèmes de détection ou de reconnaissance de forme. Nous considérons que la perception d'une forme particulière de relief est contingente de la perception de saillances, définies ici comme des points qualitativement remarquables. La description d'une silhouette inclut les relations topologiques qui relient les formes de relief détectées. Afin de tenir compte de l'imprécision des frontières des formes de relief, nous proposons une extension de la méthode 9-intersection. Les relations méréologiques, quant à elles, nous sont utiles pour dériver des représentations à différents niveaux d'abstraction.
Nous avons implanté le modèle en Java. Le prototype réealisé permet de définir des catégories, d'analyser des silhouettes, de déterminer les relations topologiques qui lient les formes détectées et d'obtenir une description à différents niveaux d'abstraction.
Cette thèse a bénéficié du soutien financier de la Région Bretagne.
Hurpeau, Jean-Christophe. "Étude de modifications de la sensibilité cutanée après microchirurgie reconstructrice des membres supérieurs et de la main en particulier : métrologie et modélisation informatique associées." Vandoeuvre-les-Nancy, INPL, 1996. http://www.theses.fr/1996INPL151N.
Full textBarthet, Mathieu. "De l'interprète à l'auditeur : une analyse acoustique et perceptive du timbre musical." Phd thesis, Aix-Marseille 2, 2008. http://tel.archives-ouvertes.fr/tel-00418296.
Full textUne étude de dissemblance réalisée sur des sons de synthèse de clarinette, obtenus à partir d'un modèle physique, a permis d'évaluer l'influence du contrôle instrumental (pression d'alimentation et pince) sur les timbres produits par l'instrument. Des enregistrements d'extraits musicaux joués un grand nombre de fois par un même clarinettiste professionnel sur un instrument naturel selon différentes intentions musicales (“scolaire" et “expressive") ont ensuite été analysés. Les mécanismes de transmission de l'expression musicale ont ainsi pu être étudiés au travers de changements de timbre, de rythme et de dynamique. Certaines variations de timbre (variations de qualité sonore au sein des notes et entre les différentes notes) sont reproduites de manière systématique par l'interprète lorsque son intention musicale est la même. La nature de ces variations change lorsque l'intention expressive change, ce qui tend à prouver que les musiciens agissent sur certaines dimensions du timbre afin de varier leur expression. Deux expériences perceptives complémentaires ont révélé que la nature des évolutions temporelles de la brillance des notes influe sur les préférences musicales des auditeurs. La qualité musicale de séquences inexpressives, produites sur des instruments de type entretenu à l'aide d'échantillonneurs, a notamment pu être améliorée de manière significative grâce à un contrôle de la brillance par filtrage dynamique.
L'ensemble de ces travaux appuie l'idée que les variations morphologiques temporelles de timbre (par ex. variations temporelles de brillance) constituent l'un des vecteurs de l'expression musicale.
Chancel, Marie. "Étude des règles de l'intégration multisensorielle en kinesthésie et de leur évolution liée à l’âge : approches psychophysiques & modélisation bayésienne." Thesis, Université Grenoble Alpes (ComUE), 2016. http://www.theses.fr/2016GREAS047/document.
Full textIn our daily live, our central nervous system uses and integrates multiples sensory information to efficiently and accurately perceive self-body and self-limb movements. An important part of the studies on this perception, called kinesthesia, address the question of the multisensory integration principles in this particular perceptive domain. This thesis brings up some answers to this question, enlightening muscle proprioception major contribution to the elaboration of multisensory kinesthetic percepts in young adults and changes occurring with age in the rules governing the integration of vision touch and muscle proprioception in kinesthesia.Revisiting a classical illusory phenomenon implicitly supposed of visual origin, the mirror paradigm, we investigate how the contralateral muscle proprioceptive afferences contribute to the visuo-proprioceptive estimation of self-arm movements. Indeed, the movement illusion of an arm, hidden behind a mirror, created by the reflection of the contralateral arm moving in this mirror seems to emerge from the integration of visual and bilateral proprioceptive feedbacks, as attested by our results. Then, in order to estimate the relative contribution and interaction of vision, touch, and muscle proprioception, we applied specific sensory stimulations on muscle proprioception, touch and/or vision. These stimulations elicit in perfectly still participants self-hand movement illusions. Combining psychophysics and Bayesian approach, we demonstrate that visual and tactile cue integration leads to an optimization of our ability to discriminate self-hand rotation velocity. Nevertheless, kinesthesia remains biased in favor of proprioceptive cues. Finally, knowing that all sensory systems decline across life span, we study the potential evolution of multisensory integration rules in kinesthesia with age. We show a profound reshaping in the weighting of the three sensory entries in older individuals, in favor of touch and vision, probably due to a relative greater impairment of muscle proprioception
Tzvetanov, Tzvetomir. "Etude psychophysique et modélisation des traitements de bas niveau sous-tendant la vision des contours des objets." Phd thesis, Université Louis Pasteur - Strasbourg I, 2003. http://tel.archives-ouvertes.fr/tel-00004179.
Full textPonce, Alvarez Adrián. "Probabilistic models for studying variability in single-neuron and neuronal ensemble activity." Thesis, Aix-Marseille 2, 2010. http://www.theses.fr/2010AIX20706.
Full textA hallmark of cortical activity is its high degree of variability. The present work focused on (i) the variability ofintervals between spikes that single neurons emit, called spike time irregularity (STI), and (ii) the variability inthe temporal evolution of the collective neuronal activity. First, I studied the STI of macaque motor corticalneurons during time estimation and movement preparation. I found that although the firing rate of the neuronstransmitted information about these processes, the STI of a neuron is not flexible and is determined by thebalance of excitatory and inhibitory inputs. These results were obtained by means of an irregularity measure thatI compared to other existing measures. Second, I analyzed the neuronal ensemble activity of severalsomatosensory and motor cortical areas of macaques during tactile discrimination. I showed that ensembleactivity can be effectively described by the Hidden Markov Model (HMM). Both sensory and decision-makingprocesses were distributed across many areas. Moreover, I showed that decision-related changes in neuronalactivity rely on a noise-driven mechanism and that the maintenance of the decision relies on transient dynamics,subtending the conversion of a decision into an action. Third, I characterized the statistics of spontaneous UP andDOWN states in the prefrontal cortex of a rat, using the HMM. I showed that state alternation is stochastic andthe activity during UP states is dynamic. Hence, variability is prominent both during active behavior andspontaneous activity and is determined by structural factors, thus rending it inherent to cortical organization andshaping the function of neural networks
Drouard, Vincent. "Localisation et suivi de visages à partir d'images et de sons : une approche Bayésienne temporelle et commumative." Thesis, Université Grenoble Alpes (ComUE), 2017. http://www.theses.fr/2017GREAM094/document.
Full textIn this thesis, we address the well-known problem of head-pose estimationin the context of human-robot interaction (HRI). We accomplish this taskin a two step approach. First, we focus on the estimation of the head pose from visual features. We design features that could represent the face under different orientations and various resolutions in the image. The resulting is a high-dimensional representation of a face from an RGB image. Inspired from [Deleforge 15] we propose to solve the head-pose estimation problem by building a link between the head-pose parameters and the high-dimensional features perceived by a camera. This link is learned using a high-to-low probabilistic regression built using probabilistic mixture of affine transformations. With respect to classic head-pose estimation methods we extend the head-pose parameters by adding some variables to take into account variety in the observations (e.g. misaligned face bounding-box), to obtain a robust method under realistic conditions. Evaluation of the methods shows that our approach achieve better results than classic regression methods and similar results thanstate of the art methods in head pose that use additional cues to estimate the head pose (e.g depth information). Secondly, we propose a temporal model by using tracker ability to combine information from both the present and the past. Our aim through this is to give a smoother estimation output, and to correct oscillations between two consecutives independent observations. The proposed approach embeds the previous regression into a temporal filtering framework. This extention is part of the family of switching dynamic models and keeps all the advantages of the mixture of affine regressions used. Overall the proposed tracker gives a more accurate and smoother estimation of the head pose on a video sequence. In addition, the proposed switching dynamic model gives better results than standard tracking models such as Kalman filter. While being applied to the head-pose estimation problem the methodology presented in this thesis is really general and can be used to solve various regression and tracking problems, e.g. we applied it to the tracking of a sound source in an image
Fillatreau, Philippe. "Localisation et modélisation tridimensionnelles pour un robot mobile autonome tout terrain." Phd thesis, Université Paul Sabatier - Toulouse III, 1994. http://tel.archives-ouvertes.fr/tel-00261834.
Full textSchatz, Thomas. "ABX-discriminability measures and applications." Thesis, Paris 6, 2016. http://www.theses.fr/2016PA066324.
Full textThis thesis constitutes an indirect contribution to the problem of modeling phonetic category acquisition in infancy. Some specific computational models of phonetic category acquisition have been proposed, but they were never tested extensively nor compared quantitatively to see whether they were really able to account for a sizable portion of the available empirical observations. In this thesis, we introduce ABX-Discriminability Measures and we develop a methodology based on these measures that allows to perform such a systematic evaluation. We demonstrate the interest of our framework by applying it to the evaluation of models for two related problems: phonetic category processing at birth and in adulthood. The next step, applying our framework to models of phonetic category acquisition, is left for future work.The interest of ABX-Discriminability Measures is not restricted to the particular problem of evaluating models of phonetic category processing in humans. We argue that their interest generalizes to the study of other signals than speech and other category structures than phonetic categories, as well as to other research fields than cognitive science, like low-resource engineering, data mining and artificial intelligence for example. To make this point, we study the properties of these measures in a general abstract framework and we detail the rationale for three broad family of potential applications: evaluating systems operating without explicit supervision in their ability to represent a category structure; providing simple computational models of behavior in discrimination tasks; providing descriptive measurements for representations of categorical data