Thesis (M.A.)--Miami University, Dept. of Psychology, 2005. / Title from first page of PDF document. Document formatted into pages; contains [1], iv, 35 p. : ill. Includes bibliographical references (p. 32-35).

Thesis (M.S.)--Indiana University, 2003. / Includes bibliographical references (leaves 56-64).

<p>One of the most important aspects of visual perception is inference of 3D shape from a 2D retinal image of the real world. The existence of several valid mapping functions from object to data makes this inverse problem ill-posed and therefore computationally difficult. In human vision, the retinal image is a 2D projection of the 3D real world. The visual system imposes certain constraints on the family of solutions in order to efficiently solve this inverse problem. This project specifically focuses on the aspect of minimization of standard deviation of all 3D angles (MSDA) for 3D perception. Our goal is to use a Deep Convolutional Neural Network based on biological principles derived from visual area V4 to solve 3D reconstruction using constrained minimization of MSDA. We conduct an experiment with novel shapes with human participants to collect data to test our model.</p>

This diploma thesis is about audience of theatre performance. It focuses on specificity of individual perception but also on perception of the audience as a unit and it examines relations between them. The thesis is based on interdisciplinary cooperation of theatre studies, cognitive science, neuroscience and psychology. It finds its goal at the topic analysis based on approach of theatre and cognitive studies. It offers a few ways, that is possible to use during the perception analysis and its specificity. In the final part it offers several ways of experimental research of the topic and its risks. Key words: theatre, spectator, perception, memory, cognitive science, psychology

Notre travail s’inscrit dans une conception fonctionnaliste de la mémoire qui se donne pour objectif de décrire le fonctionnement de la mémoire plus que sa structure. Notre assise théorique est basée sur le modèle Act-in qui postule que la mémoire est représentée par un ensemble de traces multidimensionnelles. Ces traces comporteraient l’ensemble des dimensions d’une expérience (e.g., motrices, sensorielles, émotionnelles). Pour rendre compte de l’émergence des connaissances, ce modèle repose sur deux mécanismes : l’activation et l’intégration. Notre travail avait pour objectif de spécifier les relations fonctionnelles entre activation et intégration à travers les mécanismes perceptifs et les mécanismes mnésiques. Nous avons également proposé un mécanisme d’interaction, permettant de rendre compte de la formation des traces sensorielles en mémoire. Comme le modèle Act-in conçoit les traces en mémoire comme étant multidimensionnelles, nous avons étudié les différentes relations entre mécanismes d’intégration, interaction et activation via le prisme de la multimodalité. Ce travail s’est organisé autour de deux grands axes expérimentaux, un premier tourné vers l’intégration en perception et l’autre vers l’intégration en mémoire.Dans un premier axe, nous avons introduit la notion d’interaction multimodale pour rendre compte des relations entre modalités sensorielles dans des tâches perceptives. Nous avons montré que lorsque des interactions sensorielles étaient systématiquement répétées, les performances de sujets étaient meilleures que celles observées dans un contexte multimodal dans lequel les modalités sensorielles n’entretenaient pas de lien. Nous avons mesuré à l’aide d’un outil mathématique, le Race Model, le gain multi-sensoriel dans diverses conditions. Nous avons montré que le Race Model mesurait un gain multi-sensoriel essentiellement lorsque la tâche n’impliquait pas de processus mnésiques (i.e., détection). Selon nous, cela montre que cet outil mathématique est plus sensible à l’interaction multimodale, lorsque les composants sensoriels ne sont pas intégrés en mémoire.Dans un second axe, nous avons développé un paradigme expérimental permettant de mesurer les effets d’un encodage sensoriel sur une récupération ultérieure (i.e., catégorisation, reconnaissance et rappel libre). Nous avons montré qu’une association multimodale congruente lors d’un encodage (e.g., image de chien et aboiement du chien) améliorait significativement les performances dans diverses tâches de mémoire. Selon nous, l’activation des dimensions auditives associées au sein de la trace explique cette facilitation lors de la récupération. Nous avons également montré qu’une association multimodale non congruente lors de l’encodage (e.g., image d’un chien et bruit blanc) perturbait la récupération en mémoire à long terme. Ce résultat apporte un argument supplémentaire en faveur des modèles de mémoire qui postulent que les connaissances sont avant tout sensorielles.Pour résumer, ce travail a permis dans un premier temps de proposer un mécanisme d’interaction multimodale nécessaire à l’intégration des composants sensoriels en mémoire. Dans un second temps, nous avons montré que des modalités sensorielles associées lors de l’encodage pouvait être réactivées ultérieurement et faciliter le traitement uniquement si les modalités sensorielles étaient sémantiquement congruentes. Ce résultat renforce la conception d’une mémoire organisée autour de connaissances avant tout sensorielles.Enfin, les résultats que nous obtenons suggèrent que la mémoire et la perception partagent des traitements communs car un trait perceptivement absent peut influencer un trait perceptivement présent et réciproquement. Cela nous permet d’envisager une relation horizontale entre mémoire et perception et remet en question la conception traditionnelle top-down des liens entre mémoire et perception. / Our conception of memory has the aim to describe the memory process more than the memory architecture. Our theoretical approach is based on the Act-in model, postulating that memory is represented by a multidimensional set of traces. These traces should include all dimensions of an experience (e.g., sensory-motor and emotional components). This model relies on two mechanisms (i.e., activation and integration) in order to reveal the emergence of knowledge. The aim of this work was to specify the functional relationships between activation and integration through memory and perception mechanisms. We also proposed a mechanism of interaction, to understand the formation of sensory memory traces. The model Act-in is composed by multidimensional memory traces. Thus, we studied the relationship between integration, interaction and activation through multimodality. This work was organized around two main axis. On one hand perceptual integration and on the other hand the memory integration.In the first axis, we have introduced the concept of multimodal interaction to show the relationship between sensory modalities in perceptual tasks. Our results revealed that when sensory interactions were systematically repeated, the performance of subjects were better than those observed in a context in which multimodal sensory modalities did not maintain a link. We measured using a mathematical tool, the Race Model, multi-sensory gain under various conditions. We have shown that the Race Model measured multi-sensory gain, mainly when the task did not involve memory processes (i.e., detection). This mathematical tool seems to be more sensitive to multimodal interaction, when the sensory components are not integrated into memory.In the second axis, we used an experimental paradigm to measure the effects of sensory encoding on subsequent retrieval tasks (i.e., categorization, recognition and free recall). Our results showed that a congruent multimodal association at encoding (e.g., image of dog and the dog barking) improved performance in various memory tasks. It seems that the activation of the sensory dimensions could explain the facilitation during retrieval. Moreover, an incongruent multimodal association at encoding (e.g., image of a dog and white noise) disturbed performance from retrieval in long-term memory. This result provides a further argument supporting sensory-based memory model.In summary, we propose a multimodal interaction mechanism for the integration of sensory components in memory. Secondly, we have shown that sensory modalities associated during encoding could be reactivated later and facilitate the processing only when the sensory modalities were semantically congruent. This result improves the assumption that memory is sensory-based.Finally, theses results suggest that memory and perception share common features. The reactivation of a sensory dimension can influence the perceptual dimension processing and vice versa. This allows us to consider a horizontal relationship between memory and perception. In such case, the traditional relationship between memory and perception, as assumed to be top-down, could be questioned.

Traffic signs are ubiquitous in the road environment, and are a crucial and cost-effective means of ensuring traffic safety. In Queensland, Australia, a class of temporary traffic signs is used for roadwork applications, and has recently been expanded to include Multiple-Message Signs: a novel, and as yet empirically untested system, which presents numerous advantages in terms of flexibility and economy. However, potential Human Factors problems surround aspects of the design of multiple message signs. In particular, the efficiency with which information is communication may be compromised relative to single message signs. The project involved a series of laboratory studies focussing on the behavioural requirements of temporary roadworks signs, beginning with the improvement of response characteristics of a sign used to communicate lane-closure. In the first empirical study, a proposed redesign of the Lane-Status sign based on principles of warning communication was found to be more effective than the original sign. The same sign was then examined in the context of multiple-message presentation, and significant cost s with respect to the speed and accuracy of responses relative to single-message presentation were found. A subsequent study investigated multiple-message presentation of speed limit information, which is more familiar and widely encountered than lane-closure information. Similar costs for multiple-message presentation were found. However, there were data to suggest that the additional information in multiple-message signs, if used correctly, may offer drivers and additional incentive to reduce speed. The next two studies were concerned with the effects of presentation, content type (e.g., speed-limit, symbolic and text-based) and visual clutter on recognition and visual search for traffic signs. The presentation of information via multiple message signs was compared with spatially distributed single message signs, with information matched with respect to message length and sign content. Overall advantages for multiple message presentation were found for recognition, which were particularly robust for speed limit information. The results for visual search indicated some benefits to multiple message presentation, but were less decisive than those for recognition. The final two studies investigated the possibility of improving response characteristics of single-message signs with sequential, rather then individual presentation. The role of the temporal interval between successive sign presentations was explored, which has implications for the feasibility of sequential presentation in practice. In addition, the nature of sign pairs presented in sequence – identical, or of the same meaning – was found to be important. Generally, benefits to sequential presentation diminished with increases in the temporal separation interval. In terms of practical implications, the viability of introducing the proposed design changes to the Lane Status sign is discussed. The results pertaining to multiple-message sign are evaluated in the context of published design guidelines, and theoretical and practical issues surrounding sequential sign presentation are reviewed. Finally, possibilities for future research, in part based on the limitations of the current work, are expounded.

Traffic signs are ubiquitous in the road environment, and are a crucial and cost-effective means of ensuring traffic safety. In Queensland, Australia, a class of temporary traffic signs is used for roadwork applications, and has recently been expanded to include Multiple-Message Signs: a novel, and as yet empirically untested system, which presents numerous advantages in terms of flexibility and economy. However, potential Human Factors problems surround aspects of the design of multiple message signs. In particular, the efficiency with which information is communication may be compromised relative to single message signs. The project involved a series of laboratory studies focussing on the behavioural requirements of temporary roadworks signs, beginning with the improvement of response characteristics of a sign used to communicate lane-closure. In the first empirical study, a proposed redesign of the Lane-Status sign based on principles of warning communication was found to be more effective than the original sign. The same sign was then examined in the context of multiple-message presentation, and significant cost s with respect to the speed and accuracy of responses relative to single-message presentation were found. A subsequent study investigated multiple-message presentation of speed limit information, which is more familiar and widely encountered than lane-closure information. Similar costs for multiple-message presentation were found. However, there were data to suggest that the additional information in multiple-message signs, if used correctly, may offer drivers and additional incentive to reduce speed. The next two studies were concerned with the effects of presentation, content type (e.g., speed-limit, symbolic and text-based) and visual clutter on recognition and visual search for traffic signs. The presentation of information via multiple message signs was compared with spatially distributed single message signs, with information matched with respect to message length and sign content. Overall advantages for multiple message presentation were found for recognition, which were particularly robust for speed limit information. The results for visual search indicated some benefits to multiple message presentation, but were less decisive than those for recognition. The final two studies investigated the possibility of improving response characteristics of single-message signs with sequential, rather then individual presentation. The role of the temporal interval between successive sign presentations was explored, which has implications for the feasibility of sequential presentation in practice. In addition, the nature of sign pairs presented in sequence – identical, or of the same meaning – was found to be important. Generally, benefits to sequential presentation diminished with increases in the temporal separation interval. In terms of practical implications, the viability of introducing the proposed design changes to the Lane Status sign is discussed. The results pertaining to multiple-message sign are evaluated in the context of published design guidelines, and theoretical and practical issues surrounding sequential sign presentation are reviewed. Finally, possibilities for future research, in part based on the limitations of the current work, are expounded.

<p>The technology availability and hype of virtual reality has intersected as of 2016, bringing a lot of attention to virtual reality (VR) games. The research into this technology has been ongoing since its early exploration in the 1970’s. Methods have been created to try to understand and predict what causes phenomena such as motion and simulator sickness in the human computer interaction devices. Heuristics have been developed to aid the design these applications and to avoid such discomforting circumstances. Concepts and their continued exploration have been growing to improve the experience of virtual reality technology and continue the expanding imagination of what is capable with this technology. All the tools are available to create highly immersive virtual experiences, but it is up to the interpretant, the user to decide how believable, immersive and enjoyable these experiences are. By conducting an interview study on a part of the VR gaming community, more can be understood about the success and failure of immersive design by exploring the experiences of certain highly immersive games. This study takes the opportunity to listen to gamers as a means of highlighting significant key challenges and characteristics creating immersive VR games. By evaluating transcripts via thematic analysis, themes were created to understand and categorize the various aspects that are most important to the immersion and embodiment in VR games. This study and themes drawn from it reflect on the experiences and feelings of experienced VR gamers in the context of role-playing games.</p>