A cintilografia renal é um exame amplamente utilizado para a avaliação visual do funcionamento do córtex renal. Ele permite visualizar a concentração do radiofármaco, o tamanho, a forma, a simetria e a posição dos rins. No entanto, a avaliação visual das impressões diagnósticas dessas imagens tende a ser um processo subjetivo. Isso faz com que ocorra uma significativa variabilidade entre as interpretações feitas por diferentes especialistas. Assim, este trabalho tem como objetivo propor medidas quantitativas que refletem impressões diagnósticas comumente observadas por especialistas nas imagens de cintilografia renal. São atribuídos valores numéricos a essas impressões, o que, potencialmente, reduz a subjetividade e a variabilidade da interpretação das descobertas. A fim de permitir a extração dessas medidas, um método de segmentação específico para essas imagens também é proposto. Os resultados indicam que as medidas propostas atingem níveis de concordância de no mínimo 90% dos casos quando comparadas com a avaliação visual de especialistas. Esses resultados sugerem que as medidas podem ser usadas para reduzir a subjetividade na avaliação das imagens, já que elas fornecem uma alternativa quantitativa e objetiva para reportar as impressões diagnosticas das imagens de cintilografia renal. / Renal scintigraphy is a well established functional technique for the visual evaluation of the renal cortical mass. It allows the visualization of the radiopharmaceutical tracer distribution, the size, the shape, the symmetry, and the position of the kidneys. However, the visual diagnostic impressions for these images tend to be a subjective process. It causes significant variability in the interpretation of findings. Thus, this work aims at proposing quantitative measures that reflect common diagnostic impressions for those images. These measures can potentially minimize the inter-observer variability. In order to make possible the extraction of these measures, a specific segmentation method is also proposed. The results indicate that our proposed features agree in at least 90% of the cases with the specialists visual evaluation. These results suggest that the features could be used to reduce the subjectivity in the evaluation of the images, since they provide a quantitative and objective alternative to report the diagnostic impressions.

Inclusive education has increased in popularity over the last two decades. Inclusiveclassrooms consist of children from diverse backgrounds and with variousimpairments, with special educational needs. Children with special educationalneeds in vision require assistive tools to access visual information. Topics thatmake strong use of such information, e.g. geometry, can therefore be challengingfor children with visual impairments or blindness (VIB) as well as their teachers.Assistive tools like video magniers and physical objects in the shape ofgeometric gures are used to aid VIB in learning basic geometry. Researchershave investigated tools and applications to aid VIB in learning geometry byusing haptic or auditory feedback as well as combinations of these. However,the tools are often not created with considering the classroom environment. Tounderstand the situation and challenges of these two stakeholders (VIBs andtheir teachers) better, a preliminary questionnaire was conducted. Five participantsof each of the two groups lled out the questionnaires. A concept for anAndroid tablet application to convey pre-existing geometric shapes and theirannotations to VIB, using sound feedback and sonication was created andpartly developed in two incremental implementation phases. The shapes arerecreated on the tablet using elements like circles and paths and then enhancedwith sound feedback to guide VIB to the various elements. With an iterativedesign, the two prototype increments were evaluated. The rst evaluation wasan usability study. Using seven blindfolded, young adult participants, feedbackwas gathered on dierent combinations between sound, tactile, and vibrationfeedback, as well as the general usability of the application for an validation ofthe approach. Each participants was presented the dierent approaches in anotherorder. For each approach, they were asked to explore three basic geometricshapes and then, without blindfold, draw these shapes. The time for applicationinteraction was measured and overall observation notes were taken. After eachapproach a short questionnaire was given to the participant. Another longerquestionnaire was conducted after all approaches were tested. It was foundthat sound feedback alone is not sucient in precisely conveying the geometricshapes to the users. However, combined with tactile objects to mark positionsof vertices on the screen the result improved signicantly and this approach wasfavored by the participants. The usability of the application is good but hasroom for improvement. A study was conducted with two teachers as expertslater on with the second prototype version. Two teachers with several yearsof experience teaching geometry to VIB in inclusive settings were interviewedconcerning the expected usability of the application for VIB and in the classroom,as well as general challenges and possibilities of including the applicationinto geometry lectures. As the teachers were located in Germany, they weresent a video demonstrating the application. One interview was conducted overemail while the other was conducted over Skype. The only challenges identiedare the media anity of teachers and the reliability of technology. VIB are notexpected to face challenges in utilizing the application and teachers could havemore freedom in the geometric shapes they present to their class.

This paper is based on Amouzad Mahdiraji’s research about an agent-based ambulance dispatchment simulator. A simulator which is currently not focused on visual user interfaces and usability. In this study, we implemented design science to structure our research for creating a user-friendly Graphical User Interface (GUI) prototype for an ambulance dispatchment system. We used theories concerning usability and human cognition to support the development of the prototype which was later evaluated through interviews. Cognitive overload or confusion concerning a GUI can cause stress and problems which results in various errors and costs, affecting businesses and users in multiple ways. Our study resulted in a second version of the prototype in which interview analysis and theories, mainly the Gestalt principles and heuristics, were applied. Design decisions for the second version were made to improve the usability and minimize risks of confusion. Usability in a graphical user interface can be achieved in different approaches, this study shows one of them. However, to create a suitable interface for an ambulance dispatchment system, more research is required.

De l'avènement des nouvelles technologies, du « tout numérique », de l'essor d'Internet et plus particulièrement du World Wide Web (ou web) résulte une profusion d'informations à la portée de tous. Néanmoins, la localisation des informations pertinentes au sein de cette masse informationnelle reste posé.<br />Le contexte de mes travaux est la recherche d'information textuelle sur le Web et s'inscrit dans le cadre du GDR I3 du CNRS. Ma thèse s'intitule « Interface adaptative pour l'aide à la recherche d'information sur le web ». Elle concerne la conception et la réalisation d'une interface permettant d'aider l'utilisateur dans sa démarche de recherche d'information afin qu'il puisse trouver plus efficacement des documents pertinents. Le caractère « adaptatif » de cette interface réside dans le fait qu'elle s'adapte aux besoins de l'utilisateur en lui proposant une aide personnalisée. Ce point est d'autant plus important que la vulgarisation de la recherche d'information implique la prise en compte de la différence entre les usagers et de leur spécificités. Il s'agit donc d'étudier et de proposer des outils permettant d'aider l'utilisateur dans sa tâche de recherche d'information en caractérisant notamment sa place au sein d'un tel processus. Nos travaux ont permis la conception et l'implantation d'un système nommé Easy-DOR « Easy DOcument Retrieval ». L'aide que nous proposons à l'utilisateur au travers de ce système intervient à tous les niveaux de sa recherche d'information :<br />- en amont du processus de recherche. Le système aide l'utilisateur à faire évoluer son expertise des domaines relatifs à ses centres d'intérêt afin qu'il puisse effectuer de meilleures recherches ultérieures,<br />- au cours du processus de recherche. Le système exploite les informations provenant de l'utilisateur pour tenter d'identifier ses besoins et ainsi lui apporter rapidement des documents pertinents. Par ailleurs, nous proposons une interface de visualisation lui permettant de mieux apprécier de façon globale les résultats de recherche d'information provenant d'un outil de recherche intégré,<br />- en aval du processus de recherche. Le système propose à l'utilisateur une mise à jour ainsi qu'une aide à l'organisation des documents pertinents qu'il souhaite mémoriser au travers de ses signets (ou favoris).<br />Par ailleurs, l'aide à la recherche d'information sur laquelle repose notre démarche est basée sur un aspect coopératif. Nous privilégions, en effet, le partage des informations pour les diffuser aux utilisateurs possédant les mêmes centres d'intérêt.

Le travail effectué lors de cette thèse concerne le développement d'un système cognitif artificiel autonome. La solution proposée repose sur l'hypothèse que la curiosité est une source de motivation d'un système cognitif dans le processus d'acquisition des nouvelles connaissances. En outre, deux types distincts de curiosité ont été identifiés conformément au système cognitif humain. Sur ce principe, une architecture cognitive à deux niveaux a été proposée. Le bas-niveau repose sur le principe de la saillance perceptive, tandis que le haut-niveau réalise l'acquisition des connaissances par l'observation et l'interaction avec l'environnement. Cette thèse apporte les contributions suivantes : A) Un état de l'art sur l'acquisition autonome de connaissance. B) L'étude, la conception et la réalisation d'un système cognitif bas-niveau basé sur le principe de la curiosité perceptive. L'approche proposée repose sur la saillance visuelle réalisée grâce au développement d'un algorithme rapide et robuste permettant la détection et l'apprentissage d'objets saillants. C) La conception d'un système cognitif haut-niveau, basé sur une approche générique, permettant l'acquisition de connaissance à partir de l'observation et de l'interaction avec son environnent (y compris avec les êtres humains). Basé sur la curiosité épistémique, le système cognitif haut-niveau développé permet à une machine (par exemple un robot) de devenir l'acteur de son propre apprentissage. Une conséquence substantielle d'un tel système est la possibilité de conférer des capacités cognitives haut-niveau multimodales à des robots pour accroître leur autonomie dans un environnement réel (environnement humain). D) La mise en œuvre de la stratégie proposée dans le cadre de la robotique autonome. Les études et les validations expérimentales réalisées ont notamment confirmé que notre approche permet d'accroître l'autonomie des robots dans un environnement réel / The work accomplished in this thesis concerns development of an autonomous machine cognition system. The proposed solution reposes on the assumption that it is the curiosity which motivates a cognitive system to acquire new knowledge. Further, two distinct kinds of curiosity are identified in conformity to human cognitive system. On this I build a two level cognitive architecture. I identify its lower level with the perceptual saliency mechanism, while the higher level performs knowledge acquisition from observation and interaction with the environment. This thesis brings the following contribution: A) Investigation of the state of the art in autonomous knowledge acquisition. B) Realization of a lower cognitive level in the ensemble of the mentioned system, which is realizing the perceptual curiosity mechanism through a novel fast, real-world robust algorithm for salient object detection and learning. C) Realization of a higher cognitive level through a general framework for knowledge acquisition from observation and interaction with the environment including humans. Based on the epistemic curiosity, the high-level cognitive system enables a machine (e.g. a robot) to be itself the actor of its learning. An important consequence of this system is the possibility to confer high level multimodal cognitive capabilities to robots to increase their autonomy in real-world environment (human environment). D) Realization of the strategy proposed in the context of autonomous robotics. The studies and experimental validations done had confirmed notably that our approach allows increasing the autonomy of robots in real-world environment

Un contenu visuel flou peut-il être efficace pour traiter de l’information émotionnelle ? Récemment, les travaux psychologiques en traitement émotionnel de l’information visuelle font état d’un lien particulier avec l’information de basses fréquences spatiales (BFS), grossière et floue mais rapide, qui permettrait la transmission très rapide de signaux au système émotionnel par rapport à une information de hautes fréquences spatiales (HFS) plus complexe et détaillée. En outre, l’information BFS serait primordiale dans la détection d’un danger potentiel de l’environnement et par conséquent envers des émotions à valence négative. Ces deux types d’informations visuelles emprunteraient des voies neuronales différentes conduisant à une segmentation de cette information visuelle dans le cerveau. Au niveau psychopathologique, des troubles neurologiques comme la maladie de Gilles de la Tourette ou encore la dépression majeure résistante sont connues pour entraîner un déficit des interactions sociales pour lesquelles les traitements émotionnels sont indispensables. Les dysfonctionnements neurologiques et psychobiologiques accompagnant ces troubles impliquent des structures spécifiques et localisées en périphéries ou enfouies dans le cerveau liées à la dichotomie fonctionnelle de l’information visuelle. Un des moyens simples pour appréhender ces traitements est la catégorisation des visages émotionnels. Cette recherche a examiné au niveau théorique et appliqué dans quelle mesure l’information visuelle autrement appelée la résolution en fréquences spatiales (FS) jouerait un rôle dans la catégorisation des expressions faciales émotionnelles (EFEs). Ainsi tant au niveau de la détection visuelle de danger qu’au niveau de l’identification des EFEs dans la maladie Gilles de la Tourette et dans la dépression majeure résistante, nous avons étudié les réponses comportementales dans les premières étapes de décryptage de l’information visuelle convoyant des indices émotionnels. Une comparaison avec une population contrôle a permis de cerner plus précisément les effets d’un filtrage en FS dans les processus de catégorisation avec la prédiction d’un bénéfice à traiter des contenus flous (BFS) par rapport à des contenus détaillés (HFS) pour des EFEs problématiques à classer en fonction de la pathologie. Nos résultats ont suggéré une meilleure identification de certaines EFEs filtrées en BFS par rapport à celles filtrées en HFS ou résolues. Nos données empiriques ont été discutées dans la perspective d’une segmentation de l’information visuelle dans le cerveau sollicitant des circuits neuronaux spécifiques favorisant l’accès de l’information visuelle aux centres émotionnels. En regard des structures cérébrales impliquées et des activités neuronales connus dans les troubles étudiées, l’activité dopaminergique des neurones sollicités pourraient expliquées en partie nos données factuelles. / What can be the efficiency of coarse scales in emotional information processing? Recently, psychological findings about emotional processing of visual information reported a particular link with low spatial frequencies (LSF), coarse and blurred but rapid, which might offer a very fast signal to emotional system compared to high spatial frequencies (HSF) more intricate and detailed. Plus, LSF information might be essential in danger detection and consequently in negative emotions classification. These two types of visual information would take different neural pathways driving to visual information segmentation in the brain. In psychopathological view, neurological disorders as Tourette syndrome or treatment-resistant depression are well known to produce social interaction troubles in which emotions are obligatory. Neurological and psychobiological dysfunctions belonging to these diseases implicate specific neural structures located at peripheral or inside the brain that are bind to functional dichotomy of visual information. One of the simplest ways to examine that processing is the categorization of emotion faces. This research has investigated according to theoretical and practical aspects the extent to which visual information or spatial frequency scaling (SF) might be implicated in categorization of emotional facial expressions (EFE). Thus, both in danger detection and EFE classification, among Tourette syndrome and treatment-resistant depression, we have studied behavioural responses during the first steps of visual information interpretation providing emotional cues. A comparison with healthy control population has given more precise effects of FS filtering in categorization processing with the hypothesis of a benefice to process coarse scales (LSF) compared to detailed signals (HSF) for the identification of difficult EFE in respect with the disorder. Our results have suggested a best identification of specific EFE filtered in LFS compared to HSF or intact images called broad spatial frequencies (BSF). Our empirical findings were argued in the perspective of visual information segmentation in the brain requesting specific neuronal circuits favouring visual information access to emotional complex. Given implicated brain areas and neuronal activities regarding studied disorders, dopaminergic innervation might explain our factual data.

Right now you are reading a sentence. Earlier, you might have been looking at a realistic picture, such as a photograph, or an outline drawing in a set of instructions. If you are a programmer, you work with sentence-like structures, such as code, or a system diagram. These are all graphic representations. To varying degrees, the effectiveness of every graphic representation relies on its ability to convey the designer’s intended meaning and elicit the intended reaction from its audience. However, the design of graphic representations, even in technical domains such as visual programming language design or interactive information visualization, currently relies heavily on general principles based solely on practice, intuition, and informal measures of effectiveness from the applied art and craft of design (as opposed to scientific analysis or theory). There is an increasing demand for a scientific understanding of design and its evaluation from stakeholders (who seek evidence for effectiveness) and designers (who seek to advance their field). Because both the creation of graphic displays and their perception are literally embodied experiences, a model was developed with an embodiment orientation, specifically based on how graphics are perceptually and cognitively processed. In my research, I found that graphic representations are constituted of two properties, pictorial and symbolic information, that emerge through two interrelated aspects of perception. In sighted individuals, for example, every graphic representation makes use of biological capabilities to process visual sensation (i.e., light hitting the retina), which are processed in relation to culturally-learned capabilities (i.e., writing). I observed how graphic representations – such as pictures, diagrams, and sentences – are “naturally selected” (i.e., during different phases of design or problem solving). From these observations, I developed a model that distinguishes and predicts the effectiveness of pictures, diagrams, and sentences, in terms of how object relations and attributes are pictorially or symbolically represented, relative to the functional roles of those representations, contexts, and in some cases, individual perceptual-cognitive differences among perceivers. This model is a step toward a science of graphics that could lead to evaluation techniques for information systems, theories for inclusive design, and ergonomically designed software programming tools.

No description available.

Right now you are reading a sentence. Earlier, you might have been looking at a realistic picture, such as a photograph, or an outline drawing in a set of instructions. If you are a programmer, you work with sentence-like structures, such as code, or a system diagram. These are all graphic representations. To varying degrees, the effectiveness of every graphic representation relies on its ability to convey the designer’s intended meaning and elicit the intended reaction from its audience. However, the design of graphic representations, even in technical domains such as visual programming language design or interactive information visualization, currently relies heavily on general principles based solely on practice, intuition, and informal measures of effectiveness from the applied art and craft of design (as opposed to scientific analysis or theory). There is an increasing demand for a scientific understanding of design and its evaluation from stakeholders (who seek evidence for effectiveness) and designers (who seek to advance their field). Because both the creation of graphic displays and their perception are literally embodied experiences, a model was developed with an embodiment orientation, specifically based on how graphics are perceptually and cognitively processed. In my research, I found that graphic representations are constituted of two properties, pictorial and symbolic information, that emerge through two interrelated aspects of perception. In sighted individuals, for example, every graphic representation makes use of biological capabilities to process visual sensation (i.e., light hitting the retina), which are processed in relation to culturally-learned capabilities (i.e., writing). I observed how graphic representations – such as pictures, diagrams, and sentences – are “naturally selected” (i.e., during different phases of design or problem solving). From these observations, I developed a model that distinguishes and predicts the effectiveness of pictures, diagrams, and sentences, in terms of how object relations and attributes are pictorially or symbolically represented, relative to the functional roles of those representations, contexts, and in some cases, individual perceptual-cognitive differences among perceivers. This model is a step toward a science of graphics that could lead to evaluation techniques for information systems, theories for inclusive design, and ergonomically designed software programming tools.

Dans ce travail de thèse, nous avons tout d’abord étudié comment l’information de mouvement est intégrée pour estimer la vitesse d’une texture aléatoire afin de la suivre réflexivement avec les yeux ou d’estimer perception son déplacement. Dans une seconde série d’études, nous avons étudié comment la distribution spatiale du contraste dans une texture affecte à la fois les mouvements oculaires de fixation et la reconnaissance perceptive. A ces fins, nous avons utilisé un nouvel ensemble de stimuli visuels, des textures pseudo-naturelles dans lesquelles on peut finement contrôler la statistique (moyenne, variance) des fréquences spatiales et/ou temporelles. La première partie explore l’intégration et le décodage de information fréquentielle spatio-temporelle visuelle pour les réponses de poursuite réflexes et la discrimination perceptive. Nous montrons que l’action tire complètement partie de la richesse du stimuli en intégrant sur toute la distribution pour estimer la vitesse : accélération initiale, précision et robustesse sont améliorées. Au contraire, la performance perceptive décroit pour des stimuli à bandes passantes larges. Cette dissociation se maintient sur une large place d’intégration temporelle. La seconde partie élargie cette approche à la distribution spatiale de l’information et à ces différentes échelles. Nous montrons que le comportement oculaire de fixation dépend de la composition fréquentielle d’une texture, en termes de moyenne et de fréquence. Saccade et micro-saccades se distribuent au cours de la fixation de façon coordonnée en fonction de cette statistique. In fine, cette dernière joue sur la carte de salience calculée à partir de l’image. / We focused on the impact of the statistical distributions of visual information on these various behavioral responses. We asked first how motion information is integrated to estimate speed in order to perform either a speed discrimination task or to control reflexive tracking eye movements. Next, we investigated how spatial distribution in textures affects both pattern recognition and fixational eye movements. To do so, we used a set of artificial stimuli that are naturalistic textures where we can maintain a tight control on their information contents as for instance their spatio-temporal frequency bandwidth. The first studies compared speed information decoding for ocular following eye movements and perceptual speed discrimination. We found a strong dissociation where ocular following take full advantage by the enlargement of the spatio-temporal frequency bandwidth while perceptual speed discrimination is largely impaired for large bandwidth stimuli. Such dissociation remains over a large temporal integration window. We propose an adaptive gain control mechanism to explain this opposite dependencies. The second series of experimental studies investigate the properties of fixation eye movements (microsaccade and saccade) as a function of the mean and variance of the spatial frequency content of visual static textures. We show that several characteristics of fixational saccades (location, direction and amplitude) varied systematically with the distribution of spatial frequencies. The spatial distribution of the fixation zones could be best predicted from the saliency maps of the stimuli.