Dissertations / Theses on the topic 'Human Artificial interface'

<p>A Real-Time Brain-Computer Interface is a technical system classifying increased or decreased brain activity in Real-Time between different body movements, actions performed by a person. Focus in this thesis will be on testing algorithms and settings, finding the initial time interval and how increased activity in the brain can be distinguished and satisfyingly classified. The objective is letting the system give an output somewhere within 250ms of a thought of an action, which will be faster than a persons reaction time. </p><p>Algorithms in the preprocessing were Blind Signal Separation and the Fast Fourier Transform. With different frequency and time interval settings the algorithms were tested on an offline Electroencephalographic data file based on the "Ten Twenty" Electrode Application System, classified using an Artificial Neural Network. </p><p>A satisfying time interval could be found between 125-250ms, but more research is needed to investigate that specific interval. A reduction in frequency resulted in a lack of samples in the sample window preventing the algorithms from working properly. A high frequency is therefore proposed to help keeping the sample window small in the time domain. Blind Signal Separation together with the Fast Fourier Transform had problems finding appropriate correlation using the Ten-Twenty Electrode Application System. Electrodes should be placed more selectively at the parietal lobe, in case of requiring motor responses.</p>

Purpose – Artificial Intelligence has seen important growth in the digital area in recent years. Our aim is to explore possible guidelines that make use of AI advances to design good user experiences for digital products. Method – The proposed methods to gather the necessary qualitative data to support our claim involve open-ended interviews with UX/UI Designers working in the industry, in order to gain a deeper understanding of their thoughts and experiences. In addition, a literature review is conducted to identify the knowledge gap and build the base of our new theory. Findings – Our findings suggest a need to embrace new technological developments in favour of enhancing UX designers’ workflow. Additionally, basic AI and ML knowledge is needed to utilise these capabilities to their full potential. Indeed, a crucial area of impact where AI can augment a designer’s reach is personalization. Together with smart algorithms, designers may target their creations to specific user needs and demands. UX designers even have the opportunity for innovation as mundane tasks are automated by intelligent assistants, which broadens the possibility of acquiring further skills to enhance their work. One result, that is both innovative and unexpected, is the notion that AI and ML can augment a designer’s creativity by taking over mundane tasks, as well as, providing assistance with certain graphics and inputs. Implications – These results indicate that AI and ML may potentially impact the UX industry in a positive manner, as long as designers make use of the technology for the benefit of the user in true human-centred practice. Limitations – Nevertheless, our study presents its own unique limitations due to the scope and time frame of this dissertation, we are bound to the knowledge gathered from a small sample of professionals in Sweden. Presented guidelines are a suggestion based on our research and not a definitive workflow.

The topic of this thesis is learning through social interaction, consisting of experiments that focus on word acquisition through imitation, and a formalism aiming to provide stronger theoretical foundations. The formalism is designed to encompass essentially any situation where a learner tries to figure out what a teacher wants it to do by interaction or observation. It groups learners that are interpreting a broad range of information sources under the same theoretical framework. A teachers demonstration, it's eye gaze during a reproduction attempt and a teacher speech comment are all treated as the same type of information source. They can all tell the imitator what the demonstrator wants it to do, and they need to be interpreted in some way. By including them all under the same framework, the formalism can describe any agent that is trying to figure out what a human wants it to do. This allows us to see parallels between existing research, and it provides a framing that makes new avenues of research visible. The concept of informed preferences is introduced to deal with cases such as "the teacher would like the learner to perform an action, but if it knew the consequences of that action, would prefer another action" or "the teacher is very happy with the end result after the learner has cleaned the apartment, but if it knew that the cleaning produced a lot of noise that disturbed the neighbors, it would not like the cleaning strategy". The success of a learner is judged according to the informed teachers opinion of what would be best for the uninformed version. A series of simplified setups are also introduced showing how a toy world setup can be reduced to a crisply defined inference problem with a mathematically defined success criteria (any learner architecture-setup pair has a numerical success value). An example experiment is presented where a learner is concurrently estimating the task and what the evaluative comments of a teacher means. This experiment shows how the ideas of learning to interpret information sources can be used in practice. The first of the learning from demonstration experiments presented investigates a learner, specifically an imitator, that can learn an unknown number of tasks from unlabeled demonstrations. The imitator has access to a set of demonstrations, but it must infer the number of tasks and determine what demonstration is of what task (there are no symbols or labels attached to the demonstrations). The demonstrator is attempting to teach the imitator a rule where the task to perform is dependent on the 2D position of an object. The objects 2D position is set at a random location within four different, well separated, rectangles, each location indicating that a specific task should be performed. Three different coordinate systems were available, and each task was defined in one of them (for example ''move the hand to the object and then draw a circle around it"). To deal with this setup, a local version of Gaussian Mixture Regression (GMR) was used called Incremental Local Online Gaussian Mixture Regression (ILO-GMR). A small and fixed number of gaussians are fitted to local data, informs policy, and then new local points are gathered. Three other experiments extends the types of contexts to include the actions of another human, making the investigation of language learning possible (a word is learnt by imitating how the demonstrator responds to someone uttering the word). The robot is presented with a setup containing two humans, one demonstrator (who performs hand movements), and an interactant (who might perform some form of communicative act). The interactants behavior is treated as part of the context and the demonstrators behavior is assumed to be an appropriate response to this extended context. Two experiments explore the simultaneous learning of linguistic and non linguistic tasks (one demonstration could show the appropriate response to an interactant speech utterance and another demonstration could show the appropriate response to an object position). The imitator is not given access to any symbolic information about what word or hand sign was spoken, and must infer how many words where spoken, how many times linguistic information was present, and what demonstrations where responses to what word. Another experiment explores more advanced types of linguistic conventions and demonstrator actions (simple word order grammar in interactant communicative acts, and the imitation of internal cognitive operations performed by the demonstrator as a response). Since a single general imitation learning mechanism can deal with the acquisition of all the different types of tasks, it opens up the possibility that there might not be a need for a separate language acquisition system. Being able to learn a language is certainly very useful when growing up in a linguistic community, but this selection pressure can not be used to explain how the linguistic community arose in the first place. It will be argued that a general imitation learning mechanism is both useful in the absence of language, and will result in language given certain conditions such as shared intentionality and the ability to infer the intentions and mental states of others (all of which can be useful to develop in the absence of language). It will be argued that the general tendency to adopt normative rules is a central ingredient for language (not sufficient, and not necessary while adopting an already established language, but certainly very conducive for a community establishing linguistic conventions).

En informatique ambiante, les objectifs de l'utilisateur peuvent émerger opportunément. Il devient, dès lors, nécessaire de générer à la volée des systèmes interactifs. Un système interactif est composé d'un noyau fonctionnel et d'une Interface Homme-Machine (IHM). Cette thèse traite de la composition d'IHM pour un objectif utilisateur et un contexte d'usage (utilisateur, plate-forme, environnement) donnés. Elle en propose un espace problème fondé sur les exigences utilisateur recueillies par une étude qualitative. Un état de l'art positionne notre travail et en montre la complémentarité par rapport aux travaux existants : la composition du modèle de tâches. La composition de l'IHM concrète est déléguée à une boîte à outils d'interacteurs définis au niveau tâches. La composition du modèle de tâches se fait par planification automatique. L'étude montre que les planificateurs existants répondent partiellement au problème. Aussi, un planificateur a été spécifiquement développé pour l'IHM. Son utilisation est illustrée dans un prototype Compose. Le travail est original à deux titres : d'une part, son approche " Composition de modèles de tâches " est une extension de la littérature ; d'autre part, la composition d'IHM est un nouveau cadre applicatif pour les algorithmes de planification.

Dans le domaine du "calcul à l'aide de mots" (CW : Computing with words), les approches linguistiques floues ont démontré leur pertinence dans de nombreux problèmes de prise de décision. En effet, elles permettent de modéliser le raisonnement humain en remplaçant les mots, les évaluations, les préférences, les choix, les souhaits, etc. par des variables ad hoc, telles que les sous-ensembles flous ou des variables plus complexes. Dans cette thèse, nous partons d'un problème concret en géolocalisation : la configuration des boîtiers permettant le suivi des mobiles à surveiller et la mise en place des alertes liées au suivi. Il s'agit de mettre en place un système offrant la possibilité de passer des objectifs métiers de l'utilisateur final exprimés sous la forme de paramètres linguistiques (grâce à un dialogue en langage naturel) à une combinaison appropriée des paramètres techniques de l'application. La recherche a visé à montrer comment on peut améliorer une interface utilisateur existante pour passer d'un traitement quantitatif à un traitement qualitatif des occurrences d'événements et des contraintes de géolocalisation. Nous avons ainsi défini les extensions théoriques qui semblaient nécessaires dans le CW : un modèle, fondé sur les 2-tuples sémantiques que nous introduisons, permet de représenter, avec une grande précision et une grande justesse, des ensembles de termes linguistiques même lorsque ces derniers sont positionnés de façon fortement déséquilibrée sur leur axe. Ces 2-tuples sémantiques ont été mis en oeuvre pour interpréter sémantiquement les termes linguistiques issus du dialogue, en leur rattachant une sémantique floue contextuelle.

The design of robotic systems is currently facing human-inspired solutions as a road to replicate the human ability and flexibility in performing motor tasks. Especially for control and teleoperation purposes, the human-in-the-loop approach is a key element within the framework know as Human-Robot Interface. This thesis reports the research activity carried out for the design of Human-Robot Interfaces based on the detection of human motion intentions from surface electromyography. The main goal was to investigate intuitive and natural control solutions for the teleoperation of both robotic hands during grasping tasks and wearable devices during elbow assistive applications. The design solutions are based on the human motor control principles and surface electromyography interpretation, which are reviewed with emphasis on the concept of synergies. The electromyography based control strategies for the robotic hand grasping and the wearable device assistance are also reviewed. The contribution of this research for the control of artificial hands rely on the integration of different levels of the motor control synergistic organization, and on the combination of proportional control and machine learning approaches under the guideline of user-centred intuitiveness in the Human-Robot Interface design specifications. From the side of the wearable devices, the control of a novel upper limb assistive device based on the Twisted String Actuation concept is faced. The contribution regards the assistance of the elbow during load lifting tasks, exploring a simplification in the use of the surface electromyography within the design of the Human-Robot Interface. The aim is to work around complex subject-dependent algorithm calibrations required by joint torque estimation methods.

Les tâches dans les environnements virtuels immersifs sont associées à des techniques et à des dispositifs d'interaction 3D (e.g. la sélection d'objets 3D à l'aide de la main virtuelle via un flystick). Alors que les environnements et les tâches deviennent de plus en plus complexes, les techniques ne peuvent plus être les mêmes pour chaque application, voire pour les différentes situations au sein d'une application. Une solution est d'adapter l'interaction en fonction des besoins de la situation pour améliorer l'utilisabilité. Ces adaptations peuvent être effectuées manuellement par le concepteur ou l'utilisateur, ou automatiquement par le système créant ainsi une interaction adaptative. La formalisation d'une telle assistance automatique nécessite la gestion d'informations pertinentes au vu de la situation. L'ensemble de ces informations fait émerger le contexte de l'interaction. L'assistance adaptative obtenue en raisonnant à partir de ces informations est ainsi consciente du contexte. De nombreuses possibilités existent pour l'obtenir. Notre objectif est une gestion du contexte qui préserve ses degrés élevés d'expressivité et d'évolutivité tout en étant facile à intégrer. Nous proposons une modélisation de ce problème par des graphes conceptuels basés sur une ontologie et gérés par un moteur externe en logique du premier ordre. Le moteur est générique et utilise une base de connaissance contenant des faits et des règles, qui peuvent être changés dynamiquement. Nous avons intégré une notion de confiance, afin d'établir l'adéquation d'une situation à la base de connaissances. La confiance des réactions est comparée à leur impact afin de ne garder que les pertinentes tout en évitant de saturer l'utilisateur. Les applications utilisent des outils qui peuvent être contrôlés par le moteur. Des capteurs permettent d'extraire des informations sémantiques pour le contexte. Des effecteurs permettent d'agir sur l'application et d'obtenir des adaptations. Un jeu d'outils et une base de connaissance pour l'interaction 3D ont été créés. De nombreuses étapes sont introduites dans la base de connaissance pour de bonnes combinaisons et une réflexion indépendante d'outils spécifiques. Nos premières applications illustrent la compréhension de la situation, dont les intérêts et difficultés de l'utilisateur, et le déclenchement d'assistances adaptées. Une étude hors ligne montre ensuite l'accès et l'évolution des étapes du moteur selon la situation. Le raisonnement sémantique générique obtenu est alors expressif, compréhensif, extensif et modifiable dynamiquement. Pour l'interaction 3D, il permet une assistance universelle automatique, ponctuelle ou manuelle à l'utilisateur et des analyses hors-lignes d'activités ou de conceptions pour le concepteur.

Nous nous positionnons dans un contexte d'informatique ambiante dans lequel il arrive que les besoins de l'utilisateur n'aient pas été prévus, notamment en situation exceptionnelle. Dans ce cas, il peut ne pas exister de système préconçu qui réponde exactement à ces besoins. Pour les satisfaire, il faut alors pouvoir composer les systèmes disponibles dans l'environnement, et le système composé doit permettre à l'utilisateur de faire des choix à l'exécution. Ainsi, l'utilisateur a la possibilité d'adapter l'exécution de la composition à son contexte. Cela signifie que la composition intègre des structures de contrôle de l'exécution, destinées à l'utilisateur : la composition est dite flexible. Dans cette thèse, nous proposons de répondre au problème de la composition flexible en contexte d'intelligence ambiante avec un planificateur produisant des plans flexibles. Dans un premier temps, nous proposons une modélisation de la planification flexible. Pour cela, nous définissons les opérateurs de séquence et d'alternative, utilisés pour caractériser les plans flexibles. Nous définissons deux autres opérateurs au moyen de la séquence et de l'alternative : l'entrelacement et l'itération. Nous nous référons à ce cadre théorique pour délimiter la flexibilité traitée par notre planificateur Lambda-Graphplan. L'originalité de Lambda-Graphplan est de produire des itérations en s'appuyant sur une approche par graphe de planification. Nous montrons notamment que Lambda-Graphplan est très performant avec les domaines se prêtant à la construction de structures itératives.

Orientador: Heloisa Vieira da Rocha<br>Dissertação (mestrado) - Universidade Estadual de Campinas, Instituto de Computação<br>Made available in DSpace on 2018-08-03T16:14:44Z (GMT). No. of bitstreams: 1 Lachi_RicardoLuis_M.pdf: 2882929 bytes, checksum: 92ae4a814aace49988e977a2ab006588 (MD5) Previous issue date: 2003<br>Resumo: A educação à distância se caracteriza pelo não existência, parcial ou total, do elemento 'presencial' na relação de aprendizagem professor-aluno. Isto é, o professor e aluno encontram-se em localidades distintas (fora de uma sala de aula) e necessitam do uso de um canal de comunicação para mediar a interação. Nos últimos anos, a educação à distância, via ambientes que utilizam a Internet como meio de comunicação, tem aumentado significativamente. Esses ambientes caracterizam-se por disponibilizar inúmeras ferramentas que dão suporte ao processo de comunicação via rede (ferramentas de bate-papo, correio eletrônico, grupos de discussão). Na literatura existem inúmeros artigos que relatam o enorme volume de trabalho necessário por parte de um professor de um curso à distância para poder acompanhar adequadamente o desenvolvimento dos alunos. Com o intuito de auxiliar o professor nessa tarefa de acompanhar o desenvolvimento dos alunos de um curso à distância foi desenvolvido o agente de interface Chapa descrito em detalhes nesta dissertação. O foco do Chapa é ajudar na análise das contribuições dos participantes de uma sessão de bate-papo ocorrida no ambiente de Educação à Distância TelEduc. Nesse sentido, o Chapa é capaz de selecionar quais as mensagens mais interessantes a um determinado usuário permitindo a este analisar somente uma fração do número total de mensagens existentes em uma sessão. Essa seleção é feita com base em um dicionário de dados específico, que constrói para cada usuário, por meio das técnicas de aprendizagem inerentes ao conceito de agentes de interface. Neste trabalho, já são apresentados alguns testes preliminares feitos com o Chapa por um grupo de usuários. No entanto, mesmo tendo sido bastante positivos os resultados desses testes iniciais, ainda é necessária a realização de outros mais, em situações reais de um curso à distância, para se validar melhor o Chapa<br>Abstract: Distance education is characterized by the partial or total absence of the "presential" element in the teacher-student relationship, that is, teacher and student are far ftom each other and need to use a communication channel to media te the interaction. In the last few years, distance education based on environments that use Internet as a communication channel has grown larger and larger. These environments make available many tools that support communication through the Internet (for example: chat, e-mail, newsgroup). In the literature, many articles relate the large amount of work needed by the teacher of a distance course to accompany properly the development of the students. In order to help the teacher with the task of accompanying the development of the students in a distance course we developed the interface agent Chapa, described in details in this dissertation. Chapa's focus is on helping the analysis of participants' contributions to a chat session that took place in T elEduc, a supporting environment for distance education. In order to do that, it is capable of selecting which messages are more interesting to one user, allowing him to analyze only a ftaction of the total number of messages that exist in a session. lt is able to do this selection based on a specific data dictionary which is built for each user by means of learning techniques inherent to the technology of interface agents. In this work, some preliminary tests were done with Chapa. However, even though the results were very positives, it is still necessary to realize more tests in real world courses in order to better proofChapa's effectiveness<br>Mestrado<br>Ciência da Computação<br>Mestre em Ciência da Computação

With improvements in actuation technology and sensory systems, it is becoming increasingly feasible to create powered exoskeletal garments that can assist with the movement of human limbs. This class of robotics referred to as human-machine interfaces will one day be used for the rehabilitation of paralysed, damaged or weak upper and lower extremities. The focus of this project was the development of an exoskeletal interface for the rehabilitation of the hands. A novel sensor was designed for use in such a device. The sensor uses simple optical mechanisms centred on a spring to measure force and position simultaneously. In addition, the sensor introduces an elastic element between the actuator and its corresponding hand joint. This will allow series elastic actuation (SEA) to improve control and safely of the system. The Hand Rehabilitation Device requires multiple actuators. To stay within volume and weight constraints, it is therefore imperative to reduce the size, mass and efficiency of each actuator without losing power. A method was devised that allows small efficient actuating subunits to work together and produce a combined collective output. This work summation method was successfully implemented with Shape Memory Alloy (SMA) based actuators. The actuation, sensory, control system and human-machine interface concepts proposed were evaluated together using a single-joint electromechanical harness. This experimental setup was used with volunteer subjects to assess the potentials of a full-hand device to be used for therapy, assessment and function of the hand. The Rehabilitation Glove aims to bring significant new benefits for improving hand function, an important aspect of human independence. Furthermore, the developments in this project may one day be used for other parts of the body helping bring human-machine interface technology into the fields of rehabilitation and therapy.

Doctor of Philosophy (PhD)<br>With improvements in actuation technology and sensory systems, it is becoming increasingly feasible to create powered exoskeletal garments that can assist with the movement of human limbs. This class of robotics referred to as human-machine interfaces will one day be used for the rehabilitation of paralysed, damaged or weak upper and lower extremities. The focus of this project was the development of an exoskeletal interface for the rehabilitation of the hands. A novel sensor was designed for use in such a device. The sensor uses simple optical mechanisms centred on a spring to measure force and position simultaneously. In addition, the sensor introduces an elastic element between the actuator and its corresponding hand joint. This will allow series elastic actuation (SEA) to improve control and safely of the system. The Hand Rehabilitation Device requires multiple actuators. To stay within volume and weight constraints, it is therefore imperative to reduce the size, mass and efficiency of each actuator without losing power. A method was devised that allows small efficient actuating subunits to work together and produce a combined collective output. This work summation method was successfully implemented with Shape Memory Alloy (SMA) based actuators. The actuation, sensory, control system and human-machine interface concepts proposed were evaluated together using a single-joint electromechanical harness. This experimental setup was used with volunteer subjects to assess the potentials of a full-hand device to be used for therapy, assessment and function of the hand. The Rehabilitation Glove aims to bring significant new benefits for improving hand function, an important aspect of human independence. Furthermore, the developments in this project may one day be used for other parts of the body helping bring human-machine interface technology into the fields of rehabilitation and therapy.

Robots and virtual agents grow rapidly in behavioural sophistication and complexity. They become better learners and teachers, cooperators and communicators, workers and companions. These artefacts – whose behaviours are not always readily understood by human intuition nor comprehensibly explained in terms of mechanism – will have to interact socially. Moving beyond artificial rational systems to artificial social systems means having to engage with fundamental questions about agenthood, sociality, intelligence, and the relationship between mind and body. It also means having to revise our theories about these things in the course of continuously assessing the social sufficiency of existing artificial social agents. The present thesis presents an empirical study investigating the social influence of physical versus virtual embodiment on people's decisions in the context of a bargaining task. The results indicate that agent embodiment did not affect the social influence of the agent or the extent to which it was perceived as a social actor. However, participants' perception of the agent as a social actor did influence their decisions. This suggests that experimental results from studies comparing different robot embodiments should not be over-generalised beyond the particular task domain in which the studied interactions took place.

L'amputation du membre supérieur, dont la prévalence est comparable à celle des maladies orphelines, induit chez les patients une perte considérable d'autonomie dans la majorité des tâches simples de la vie quotidienne. Pour pallier ces difficultés, les prothèses myoélectriques actuelles proposent une multitude de mouvements possibles. Cependant, leur contrôle non intuitif et lourd cognitivement requiert un apprentissage long et difficile, qui pousse une proportion importante de patients amputés à l'abandon de la prothèse. Dans cette thèse, nous avons cherché à identifier l'origine des difficultés et les manques du contrôle myoélectrique en comparaison au contrôle sensorimoteur naturel, dans le but à terme de proposer de meilleures solutions de restitution et de suppléance. Pour cela, nous avons manipulé diverses conditions expérimentales dans un contexte d'interface homme-machine simplifié où des sujets non amputés contrôlent un curseur sur un écran à partir de contractions isométriques, i.e. des contractions qui n'engendrent pas de mouvement. Cette condition isométrique nous a permis de nous approcher de la condition de la personne amputée contrôlant sa prothèse à partir de l'activité électrique (EMG) de ses muscles résiduels, en absence de mouvement articulaire. Durant une tâche d'atteinte de cible, nous avons entre autre démontré le bénéfice d'une adaptation conjointe du décodeur qui traduit les activités EMG en mouvement du curseur, venant s'ajouter à la propre adaptation du plan de mouvement des sujets en réponse à des perturbations orientées. De plus, il a été mis en évidence que ce bénéfice est d'autant plus important que la dynamique d'adaptation artificielle du décodeur s'inspire de celle de l'Homme. Dans des tâches d'acquisition et de poursuite de cible, impliquant davantage les mécanismes de régulation en ligne du mouvement, nous avons mis en évidence l'importance d'une congruence immédiate entre les informations sensorimotrices et la position du curseur à l'écran pour permettre des corrections rapides et efficaces. Dans une condition où le niveau de bruit du système est relativement faible, comme avec l'utilisation du signal de forces plus stable que l'habituel signal EMG, cette congruence explique, en partie, la supériorité d'un contrôle d'ordre 0 (i.e. position) sur un contrôle d'ordre 1 (i.e.} vitesse). Cependant, dès lors que le niveau de bruit est trop important, ce qui est le cas avec le signal EMG, le filtrage induit par l'intégration nécessaire au contrôle vitesse fait que celui-ci devient plus performant que le contrôle position. L'ensemble de ces résultats suggèrent qu'un décodeur adaptatif et intuitif, respectant et suppléant au mieux les boucles du contrôle sensorimoteur naturel, est le plus à même de faciliter le contrôle des futures prothèses<br>Upper limb amputation, although quite rare, induces enormous loss of autonomy for patients in most daily life activities. To overcome this loss, current myoelectric prosthesis offers a multitude of possible movements. However, current controls of these movements are typically non-intuitive and cognitively demanding, leading to a high abandon rate in response to the long and tedious learning involved. In this thesis, we aimed at identifying difficulties and gaps associated with myoelectric controls when compared to natural sensorimotor control, with the long term goal of informing the design of better solutions for prosthesis control. To do so, we manipulated several experimental conditions in a simplified human-machine interface, where non-amputated subjects controlled a cursor on a computer screen from isometric contractions, i.e. muscle contractions produced in the absence of joint movement. This isometric condition was designed to get closer to a situation in which an amputee controls a myoelectric prosthesis using electrical activity (EMG) of his/her residual muscles, without movement of the missing limb. During aiming movements, we demonstrated the benefits of adapting the decoder that translate muscle activities into cursor movement in conjunction with the own subject’s adaptation of the planned movement direction in response to oriented perturbations. Furthermore, these benefits were showed to be even more important as the artificial decoder adaptation was inspired by the modeled adaptation of a human. In reaching and tracking movements toward fixed and moving targets, which increasingly involve online movement regulations, we revealed the importance of an immediate congruency between sensorimotor information and the cursor position on the screen for timely and efficient corrections. For conditions in which the level of noise associated with the control signal is relatively low, such as when using force that is more stable than the usual EMG signal used, this congruency partly explains the better performance obtained with zero order control (i.e. position) when compared to first order control (i.e. velocity). However, when the noise level increases, as is the case with EMG signals, the filtering property associated with the integration involved in a velocity control elicits better performances than with a position control. Taken together, these results suggest that intuitive and adaptive decoder, that supplies and judiciously complements natural sensorimotor feedback loops, is promising to facilitate future prosthesis controls

Orientador: Beatriz Mascia Daltrini<br>Tese (doutorado) - Universidade Estadual de Campinas, Faculdade de Engenharia Eletrica e de Computação<br>Resumo: Obter sucesso no projeto de interface de usuário requer coordenação de diversa disciplinas para criar um produto de software de alta usabilidade.Esta é a motivação deste trabalho: definir a arquitetura de um framework que permita integrar o trabalho das diversas disciplinas durante o processo de projeto. O objetivo obter um projeto de sistema interativo onde diversos usuários possam trabalho considerando um aspecto específico da interface. Neste trabalho de tese, define-se um framework para o projeto conceitual e detalhado de sistemas interativos, o qual seque a abordagem multi-agentes na sua arquitetura. Ele possui as características multi-disciplinares tais como cooperação e flexibilidade, próprias do processo de desenvolvimento de interfaces. Para conseguir que várias pessoas trabalhem com vários agentes, determinamos que o projeto da interface seja tratado também como um conjunto de agentes, chamados de agentes da interface da aplicação (AIA). Os AIAs reúnem, na sua arquitetura, as descrições declarativas de todas as partes do projeto de uma aplicação interativa (apresentação da interface, descrição das tarefas da interface, descrição do diálogo, modelos de funções e dados) naturalmente orientada ao projetista<br>Made available in DSpace on 2018-07-22T18:10:32Z (GMT). No. of bitstreams: 1 Peixoto_CeciliaInesSosaArias_D.pdf: 7983928 bytes, checksum: 8b47016c3bb9987a5e47016cce0ff7ec (MD5) Previous issue date: 1997<br>Abstract: Attaining success in user interface design requires coordination of several subjects to create a software product with high usability. This is the motivation of this work: to define the architecture of a framework which allows integration of the work of the various subjects during the design process. The goal is to obtain an interactive system design where several users can work considering only one specific aspect of the interface design. In this the framework for conceptual and detailed design of interactive system is defined. This framework has the typical multidisciplinary characteristics of the interface design process, such as cooperation and flexibility. In order to enable different people to work with various agents in the called application interface agents (AIA). The AIAs include, in their architecture, the declarative description of all parts of the design of an interactive application (presentation of the interface, dialog description, interface task description, data and functions model) in a natural, designer-oriented way<br>Doutorado<br>Doutor em Engenharia Elétrica

L'interaction Homme-machine s'articule autour de plusieurs points-clés : établissement et maintien de l'échange, compréhension des signaux et réaction. Mes travaux de recherche s'inscrivent dans le thème de l'Interaction/Communication/Relation Homme-Machine avec pour fil conducteur l'aide à la personne en situation de handicap. Au cours de mes projets, j'ai traité la problématique de l'aide aux personnes en situation de handicap suivant quatre axes : * la réalisation d'aide technique dans le cadre de la communication écrite (interface de saisie de textes), * l'interaction émotionnelle (modèle d'affect computationnel), * la stimulation cognitive, * le robot compagnon ou compagnon artificiel Ce mémoire aborde également le thème pluridisciplinaire de l'acceptabilité et de l'intercompréhension dans les interactions et l'évaluation de l'acceptabilité d'un robot comme partenaire d'interaction par l'homme.

E-sports is currently estimated to be a billion dollar industry which is only growing in size from year to year. However the cinematography of spectated games leaves much to be desired. In most cases, the spectator either gets to control their own freely-moving camera or they get to see the view that a specific player sees. This thesis presents a system for the generation of cinematically-pleasing views for spectating real-time graphics applications. A custom real-time engine has been built to demonstrate the effect of this system on several different game modes with varying visual cinematic constraints, such as the rule of thirds. To create the cinematic views, we encode cinematic rules as cost functions that are fed into a non-linear least squares solver. These cost functions rely on the geometry of the scene, minimizing residuals based on the 3D positions and 2D reprojections of the geometry. The final cinematic view is found by altering camera position and angle until a local minimum is met. The system was evaluated by comparing video output from a traditional rigidly constrained camera and the results of our algorithm’s optimally solved views. User surveys are then used to qualitatively evaluate the system. The results of these surveys do not statistically find a preference between the cinematic views and the rigidly constrained views. In addition, we present performance and timing considerations for the system, reporting that the system can operate within modern expectations of latency when enough constraints are placed on the non-linear least squares solver.

Before there were airplanes, cars, trains, boats, or bicycles, the primary means of transportation was on foot. Unfortunately, many of the trails used by ancient travelers have long since been abandoned. We present a software tool which can help visualize and predict where these forgotten trails might lie through the use of a human-centered cost metric. By comparing the paths generated by our software with known historical trails, we demonstrate how the tool can indicate likely trails used by ancient travelers. In addition, this new tool provides novel visualizations to better help the user understand alternate paths, effect of terrain, and nearby areas of interest. Such a tool could be used by archaeologists and historians to better visualize and understand the terrain and paths around sites of historical interest. This thesis is a continuation of previous work, with emphasis on the ability to generate paths which traverse several thousand kilometers. To accomplish this, various graph simplification and path approximation algorithms are explored to construct a real-time path finding algorithm. To this end, we show that it is possible to restrict the search space for a path finding algorithm while not disrupting accuracy. Combined with a multi-threaded variant of Dijkstra's shortest path algorithm, we present a tool capable of traversing the contiguous US, a dataset containing over 19 billion datapoints, in under three hours on a 2.5 Ghz dual core processor. The tool is demonstrated on several examples which show the potential archaeological and historical applicability, and provide avenues for future improvements.

Tens of thousands of years ago, humans drew sketches that we can see and identify even today. Sketches are the oldest recorded form of human communication and are still widely used. The universality of sketches supersedes that of culture and language. Despite the universal accessibility of sketches by humans, computers are unable to interpret or even correctly identify the contents of sketches drawn by humans with a practical level of accuracy. In my thesis, I demonstrate that the accuracy of existing sketch recognition techniques can be improved by optimizing the classification criteria. Current techniques classify a 20,000 sketch crowd-sourced dataset with 56% accuracy. I classify the same dataset with 52% accuracy, but identify factors that have the greatest effect on the accuracy. The ability for computers to identify human sketches would be useful particularly in pictionary-like games and other kinds of human-computer interaction; the concepts from sketch recognition could be extended to other kinds of object recognition.

Esta tesis está enfocada en la problemática asociada a la personalización de información para usuarios que interactúan con grandes volúmenes de información. En concreto, la tesis se centra en el estudio de los sistemas de recomendación conversacionales basados en críticas como herramientas para ayudar a los usuarios en la búsqueda de productos o servicios. Los sistemas de recomendación basados en críticas permiten recoger información asociada a las preferencias del usuario a través de un proceso conversacional, en el cual el usuario puede realizar una retroalimentación a las recomendaciones recibidas. La retroalimentación recibida desde el usuario, permite al recomendador adquirir conocimiento acerca de las preferencias que tiene el usuario respecto del producto que desea y generar un perfil asociado al usuario. Este tipo de técnica de recomendación en forma de diálogo cíclico, que caracteriza a los sistemas de recomendación basados en críticas, ha demostrado resultados satisfactorios en estudios previos orientados principalmente al dominio de comercio electrónico. Sin embargo, cuando un usuario en el mundo real se encuentra en una tienda física, tiene la posibilidad de interactuar y comunicarse con otros compradores. Aún más, hoy en día existen entornos virtuales colaborativos, la mayoría en tres dimensiones, donde los usuarios se pueden comunicar con otras personas y sentirse inmersos en un mundo virtual. Considerando todo lo anterior, esta tesis se focaliza en aprovechar el comportamiento colaborativo de las personas para mejorar las recomendaciones de productos o servicios. En concreto, el objetivo de la tesis es definir e integrar nuevas técnicas de personalización con interfaces de usuario inteligentes, basadas en el comportamiento colaborativo del ser humano y en la información recogida de las interacciones de las personas tanto con la interfaz como con otros usuarios. La tesis se divide en dos partes. En la primera parte, describo las nuevas técnicas de recomendación que incorporan información colaborativa. En la segunda parte, presento nuevas interfaces de usuario inteligentes implementadas en diferentes entornos de visualización y que integran las técnicas de recomendación propuestas en esta tesis. Las nuevas técnicas de recomendación que propongo integran, por un lado, opiniones de los usuarios acerca de los productos, por otro lado, integran recomendaciones históricas de otros usuarios que han utilizado el sistema de recomendación previamente y finalmente integran avances de estudios previos que han demostrado mejorar los procesos de recomendación. Adicionalmente, propongo otras técnicas basadas en información colaborativa instantánea que se puede obtener de parte del sistema de recomendación, como también de otros usuarios que están utilizando el sistema en el mismo momento. Las interfaces de usuario inteligentes que propongo se basan en entornos de visualización en 2D y 3D. En primer lugar, propongo un framework ba sado en un entorno web en dos dimensiones que permite integrar técnicas basadas en información proveniente de opiniones y recomendaciones histó ricas de otros usuarios. En segundo lugar, propongo un framework basado en un entorno virtual colaborativo en tres dimensiones que permite integrar técnicas basadas en información colaborativa que se genera por la interacción de los usuarios dentro del entorno virtual. Es necesario destacar, que ambos frameworks son modulares y escalables, permitiendo integrar diferentes algoritmos de recomendación basados en críticas. Para validar las técnicas de recomendación y las interfaces de usuario propuestas en esta tesis, defino dos metodologías de evaluación. En concreto, para la evaluación de las técnicas de recomendación utilizo un simulador que por medio de un “usuario artificial” permite evaluar el funcionamiento de las técnicas de recomendación simulando varias pruebas con usuarios reales de forma automática. Adicionalmente, para los frameworks que imple mentan las interfaces de usuario inteligentes, he realizado evaluaciones con usuarios reales. El análisis de los resultados obtenidos mediante las evaluaciones con el simulador y con los usuarios reales, revela que tanto las nuevas técnicas propuestas, como las interfaces de usuario inteligentes obtienen resultados muy satisfactorios, los cuales superan significativamente a los resultados obtenidos con técnicas de recomendación de estudios previos. Es necesario destacar, que todos los resultados han sido corroborados por métodos estadísticos. De acuerdo a los resultados obtenidos, puedo concluir que la incorporación de información colaborativa en nuevas técnicas de recomendación favorece la eficiencia y eficacia de los sistemas de recomendación basados en críticas. Además, las nuevas y novedosas interfaces de usuario inteligentes propuestas en esta tesis, representan un avance en el desarrollo de interfaces de usuario para sistemas de recomendación basados en críticas. Estas interfaces aparte de favorecer la integración de técnicas de recomendación colaborativa, mejoran la experiencia de los usuarios durante las sesiones de recomendación, lo cual es corroborado por los usuarios que participaron en los experimentos.<br>This thesis is focused on the problems associated with personalization of information for users who interact with large volumes of information. In particular, the thesis focuses on the study of critiquebased recommenders as tools for helping users in the search of products or services. The critique based recommenders allow to collect information about user’s preferences through a conversational process where the user makes critiques over the recommended product. This feedback allows the recommender to acquire knowledge about the user’s preferences about the desired product and crea tes a user profile. This type of recommendation technique based on a cyclical dialogue, which characterizes critical recommendation systems, has shown satisfactory results in previous studies oriented mainly to the domain of electronic commerce. However, when a real-world user is in a physical store, he has the ability to interact and communicate with other buyers. Moreover, nowadays there are collaborative virtual environments, most in 3D, where users can communicate with other people and feel immersed in a virtual world. Considering all mentioned above, this thesis focuses on harnessing the collaborative behavior of people to improve recommendations for products or services. Specifically, the objective of the thesis is to define and to integrate new techniques of personalization with intelligent user interfaces, based on the collaborative behavior of the human being and on the information gathered from the interactions of people with the interface and with other users. The thesis is divided into two parts. In the first part, I describe the new techniques of recommendation that incorporate collaborative information. In the second part, I present new intelligent user interfaces implemented in different visualization environments and that integrate the recommendation techniques proposed in this thesis. My proposals of the new recommendation techniques integrate, on the one hand, user opinions about the products, on the other hand, integrate his torical recommendations of other users that have used the recommendation system previously and finally integrate advances of previous studies that have shown improvements in the recommendation process. Additionally, I propose other techniques based on instant and online collaborative information that can be obtained from the recommendation system, as well as from other users that are using the system at the same time. The intelligent user interfaces that I propose are based on 2D and 3D visualization environments. First, I propose a framework based on a 2D web environment that allows the integration of techniques based on information from historical opinions and recommendations of other users. Second, I propose a framework based on a 3D collaborative virtual environment that allows the integration of techniques based on collaborative information generated by the interaction of users within the virtual environment. It is worth nothing, that both frameworks are modular and scalable, allowing to integrate different critique-based recommendation algorithms. I define two evaluation methodologies to validate the recommendation techniques and user interfaces proposed in this thesis. Specifically, for the evaluation of the recommendation techniques, I run a simulator that uses an “artificial user” that allows the evaluation of the recommendation efficiency in the proposed techniques by simulating various tests with real users automatically. In addition, I have conducted evaluations with real users for evaluating the frameworks that implement the intelligent user interfaces. The analysis of the results obtained through the simulator and real user evaluations, shows that both the proposed new techniques and the intelligent user interfaces obtain very satisfactory results, which significantly outper form the results obtained with recommendation techniques from previous studies. It is worth nothing that all the results have been corroborated by statistical methods. According to the results, I can conclude that the integration of the colla borative information in new recommendation techniques improves the efficiency and efficacy of critique-based recommenders. In addition, the new and novel intelligent user interfaces proposed in this thesis represent an ad vance in the development of user interfaces for critiquebased recommenders. These interfaces, besides facilitate the integration of collaborative recommen dation techniques, improve the users’ experience during the recommenda tion sessions, which is corroborated by the users who participated in the experiments.

Depuis les premiers travaux de Moreno en 1934, l'analyse de réseaux sociaux s'est toujours accompagnée de diagrammes représentant les relations entre individus. Depuis, la visualisation de graphes n'a cessé de se développer au sein des communautés de visualisation d'information et de dessin de graphes. Au-delà des travaux se concentrant sur les problèmes combinatoires et algorithmiques posés par les dessins de graphes on parle désormais de fouille visuelle de réseaux et plus généralement de visual analytics en intégrant l'utilisateur au coeur de l'analyse. Dans cette thèse nous nous concentrons sur les supports visuels interactifs stéréoscopiques qui permettent de piloter le processus de fouille. En utilisant un environnement expérimental ad hoc, nous essayons d'en évaluer l'impact dans une tâche très populaire de détection de structures en communautés. A travers différentes expériences, nous montrons que pour une classe de graphes particulière la 2D semble plus adapté pour les graphes simples tandis que la 3D stéréo est bénéfique pour les graphes les plus complexes. Nous identifions également des différences dans l'usage des interactions entre la mono et la stéréo, ce qui semble illustrer des différences de comportement caractérisant des différences de stratégie d'utilisation entre ces deux conditions. Nos travaux se prolongent sur le plan technologique par le développement d'une bibliothèque permettant des interactions "mains-libres" adaptées à la fouille visuelle debout devant un grand écran.

Depuis les premiers travaux de Moreno en 1934, l’analyse de réseaux sociaux s’est toujours accompagnée de diagrammes représentant les relations entre individus. Depuis, la visualisation de graphes n’a cessé de se développer au sein des communautés de visualisation d’information et de dessin de graphes. Au-delà des travaux se concentrant sur les problèmes combinatoires et algorithmiques posés par les dessins de graphes on parle désormais de fouille visuelle de réseaux et plus généralement de visual analytics en intégrant l’utilisateur au coeur de l’analyse. Dans cette thèse nous nous concentrons sur les supports visuels interactifs stéréoscopiques qui permettent de piloter le processus de fouille. En utilisant un environnement expérimental ad hoc, nous essayons d’en évaluer l’impact dans une tâche très populaire de détection de structures en communautés. A travers différentes expériences, nous montrons que pour une classe de graphes particulière la 2D semble plus adapté pour les graphes simples tandis que la 3D stéréo est bénéfique pour les graphes les plus complexes. Nous identifions également des différences dans l’usage des interactions entre la mono et la stéréo, ce qui semble illustrer des différences de comportement caractérisant des différences de stratégie d’utilisation entre ces deux conditions. Nos travaux se prolongent sur le plan technologique par le développement d’une bibliothèque permettant des interactions “mains-libres” adaptées à la fouille visuelle debout devant un grand écran<br>Ever since the pioneering work of Moreno in 1934, social network analysis has always included drawings depicting relationships between actors. From these days, the graph visualization field has grown within the graph drawing and information visualization communities. Besides the algorithmic and combinatoric questions arising from graph drawings, new challenges now include networks visual data mining. Usually referred to as visual analytics it involves the integration of the user at the heart of the analysis. In this thesis, we focus on interactive and stereoscopic visual restitutions allowing the user to drive the mining process. Using a ad hoc experimental environment, we try to assert their its impact on a popular task of community detection. Through several experiments, we show that for a specific class of graphs, 2D seems more adapted for the easier graphs while stereoscopic 3D is beneficial for the more complex ones. We also identify some differences in the interactions between the stereo and mono conditions, which seems to indicate behavioral differences emerging from differing interaction strategies. We also propose some prospects such as the implementation of a library allowing hands-free interactions adapted to visual mining in front of a large screen

We have designed a simulation that can be used to train police officers. Digital simulations are more cost-effective than a human role play. Use of force decisions are complex and made quickly, so there is a need for better training and innovative methods. Using this simulation, we are measuring the degree of presence that a human experience in a virtual environment. More presence implies better training. Participants are divided into two groups in which one group performs the experiment using a screen, keyboard, and mouse, and another uses virtual reality controls. In this experiment, we use subjective measurements and physiological measurements. We offer a questionnaire to participants before and after play. We also record the participants change in heart rate, skin conductivity and skin temperature using Empatica device. By comparing the data collected from both groups, we prove that people experience more presence in the virtual environment.

Cette thèse propose une nouvelle approche de la vision par ordinateur, insistant sur le rôle joué par les différentes informations et connaissances présentes dans un système intégré de vision. Ces connaissances peuvent être vues comme un but à atteindre dans la chaine de traitement ou comme un instrument pour guider ces différents traitement. Ce double rôle est mis en évidence en insistant sur la gestion des informations disponibles à tout moment pour effectuer la tâche de reconnaissance. Une approche distribuée est proposée, sous la forme d'un système multi-agents, où chaque agent est spécialisé pour la reconnaissance d'un concept de l'image. Le modèle générique d'agent est composé de quatre comportements de base : perception (croissance de région, suivi de contour), interaction (fusion de primitives), différenciation (interprétation de primitive) et reproduction (stratégie de focalisation des agents). Ces comportements prennent des décisions à l'aide d'évaluations multi-critères, faisant référence à plusieurs critères spécialisés. Des capacités d'adaptation et de coopération sont nécessaires aux agents pour mener à bien leurs tâches, et ce tant au niveau local qu'au niveau des groupes d'agents (travaillant sur un même groupement perceptuel). Ce système est appliqué pour la reconnaissance des cellules en mouvement. Chaque cellule est divisée en plusieurs composantes (noyau, pseudopode, cytoplasme, ...) qui servent de concepts de base pour la spécialisation des agents.

This dissertation addresses the difficulties of semantic segmentation when dealing with an extensive collection of images and 3D point clouds. Due to the ubiquity of digital cameras that help capture the world around us, as well as the advanced scanning techniques that are able to record 3D replicas of real cities, the sheer amount of visual data available presents many opportunities for both academic research and industrial applications. But the mere quantity of data also poses a tremendous challenge. In particular, the problem of distilling useful information from such a large repository of visual data has attracted ongoing interests in the fields of computer vision and data mining. Structural Semantics are fundamental to understanding both natural and man-made objects. Buildings, for example, are like languages in that they are made up of repeated structures or patterns that can be captured in images. In order to find these recurring patterns in images, I present an unsupervised frequent visual pattern mining approach that goes beyond co-location to identify spatially coherent visual patterns, regardless of their shape, size, locations and orientation. First, my approach categorizes visual items from scale-invariant image primitives with similar appearance using a suite of polynomial-time algorithms that have been designed to identify consistent structural associations among visual items, representing frequent visual patterns. After detecting repetitive image patterns, I use unsupervised and automatic segmentation of the identified patterns to generate more semantically meaningful representations. The underlying assumption is that pixels capturing the same portion of image patterns are visually consistent, while pixels that come from different backdrops are usually inconsistent. I further extend this approach to perform automatic segmentation of foreground objects from an Internet photo collection of landmark locations. New scanning technologies have successfully advanced the digital acquisition of large-scale urban landscapes. In addressing semantic segmentation and reconstruction of this data using LiDAR point clouds and geo-registered images of large-scale residential areas, I develop a complete system that simultaneously uses classification and segmentation methods to first identify different object categories and then apply category-specific reconstruction techniques to create visually pleasing and complete scene models.

Embodied Conversational Agents (ECA) form part of a range of virtual characters whose intended purpose include engaging in natural conversations with human users. While works in literature are ripe with descriptions of attempts at producing viable ECA architectures, few authors have addressed the role of episodic memory models in conversational agents. This form of memory, which provides a sense of autobiographic record-keeping in humans, has only recently been peripherally integrated into dialog management tools for ECAs. In our work, we propose to take a closer look at the shared characteristics of episodic memory models in recent examples from the field. Additionally, we propose several enhancements to these existing models through a unified episodic memory model for ECAÂ s. As part of our research into episodic memory models, we present a process for determining the prevalent contexts in the conversations obtained from the aforementioned interactions. The process presented demonstrates the use of statistical and machine learning services, as well as Natural Language Processing techniques to extract relevant snippets from conversations. Finally, mechanisms to store, retrieve, and recall episodes from previous conversations are discussed. A primary contribution of this research is in the context of contemporary memory models for conversational agents and cognitive architectures. To the best of our knowledge, this is the first attempt at providing a comparative summary of existing works. As implementations of ECAs become more complex and encompass more realistic conversation engines, we expect that episodic memory models will continue to evolve and further enhance the naturalness of conversations.<br>M.S.Cp.E.<br>School of Electrical Engineering and Computer Science<br>Engineering and Computer Science<br>Computer Engineering MSCpE

La progression rapide de la technologie a donné lieu à un panorama riche et varié de dispositifs numériques : caméra, téléphones mobiles, GPS, tablettes tactiles, liseuses numériques, robots, télévisions, éléments de domotique... La majorité de ces appareils sont aujourd'hui connectés à Internet. Et en plus de leurs fonctionnalités principales, ils permettent à leur propriétaire de rester en contact avec " le monde " à l'aide de logiciels de communication, de personnages virtuels ou de robots. Tous ces dispositifs numériques fonctionnent indépendamment les uns des autres. La question qui se pose est de savoir si ces dispositifs numériques doivent être coordonnés afin de partager certaines informations et effectuer certaines actions ensembles. Cette collaboration entre les dispositifs numériques est gérée par le Compagnon Artificiel qui est en contact permanent avec un utilisateur par les biais des divers dispositifs numériques. Une architecture modulaire ArCo permettant de mettre en place un Compagnon Artificiel a été réalisée dans le cadre de ce travail. Les dispositifs numériques sont gérés par des modules spécifiques, créés grâce à un cadriciel MICE (Machines Interaction Control in their Environment). L'utilisateur final du système peut programmer des scenarii d'interaction, qui indiquent les actions que doivent effectuer les dispositifs numériques, grâce à une interface de programmation visuelle AmbiProg. Chaque scénario est interprété par un module AmbiLive. Les conflits d'accès aux dispositifs numériques sont gérés par un module AmbiCop. Un ensemble d'évaluations a permis de valider expérimentalement l'architecture ArCo et de répondre à des problématiques d'interaction homme-machine.

La progression rapide de la technologie a donné lieu à un panorama riche et varié d’objets numériques : caméra, téléphones mobiles, GPS, tablettes tactiles, liseuses numériques, robots, télévisions, domotique. . . La majorité de ces appareils sont aujourd'hui connectés à Internet. Et en plus de leurs fonctionnalités principales, ils permettent à leur propriétaire de rester en contact avec « le monde » à l'aide de logiciels de communication, de personnages virtuels ou de robots. Tous ces dispositifs numériques fonctionnent indépendamment les uns des autres. La question qui se pose est de savoir si ces dispositifs numériques doivent être coordonnés afin de partager certaines informations et effectuer certaines actions ensembles. Cette collaboration entre les dispositifs numériques est gérée par le Compagnon Artificiel qui est en contact permanent avec un utilisateur par les biais des divers dispositifs numériques. Une architecture modulaire ArCo permettant de mettre en place un Compagnon Artificiel a été réalisée dans le cadre de ce travail. Les dispositifs numériques sont gérés par des modules spécifiques, créés grâce à un cadriciel MICE (Machines Interaction Control in their Environment). L'utilisateur final du système peut programmer des scenarii d'interaction, qui indiquent les actions que doivent effectuer les dispositifs numériques, grâce à une interface de programmation visuelle AmbiProg. Chaque scénario est interprété par un module AmbiLive. Les conflits d'accès aux dispositifs numériques sont gérés par un module AmbiCop. Un ensemble d'évaluations a permis de valider expérimentalement l'architecture ArCo et de répondre à des problématiques d'interaction homme-machine<br>The rapid advance in technology allowed manufacturers to offer a varied panorama of digital devices: camera, mobile phones, GPS, tactile tablets, digital readers, robots, televisions, home automation elements… Nowadays, most of these devices are connected to the Internet. Thus, their users can not only take benefit from their functionalities but also be in touch with “everyone” thanks to communicative software, virtual characters or robots. It raises a main question: do these digital devices have to be coordinated in order to share information and achieve tasks together? This collaboration between digital devices could be managed by an Artificial Companion which could permanently be in touch with a user throughout these devices. We realized a modular architecture called ArCo which allows the creation of an Artificial Companion. Some specialized modules can manage the digital devices. These modules are implemented with a framework called MICE (Machines Interaction Control in their Environment). The end-user can create interaction scenarios with a visual programming interface called AmbiProg. The scenarios contain actions that digital devices have to achieve. Each scenario is interpreted by the AmbiLive module. Digital devices access conflicts are managed by the AmbiCop module. Some experiments validated the ArCo architecture from an experimental point of view. They also allowed studying some Human-Machine Interaction problems

Human-computer interaction is an important area that searches for better and more comfortable systems to promote communication between humans and machines. Vision-based interfaces can offer a more natural and appealing way of communication. Moreover, it can help in the e-accessibility component of the e-inclusion. The aim is to develop a usable system, that is, the end-user must consider the use of this device effective, efficient and satisfactory. <br/>The research's main contribution is SINA, a hands-free interface based on computer vision techniques for motion impaired users. This interface does not require the user to use his upper body limbs, as only nose motion is considered. Besides the technical aspect, user's satisfaction when using an interface is a critical issue. The approach that we have adopted is to integrate usability evaluation at relevant points of the software development.

Computers have been assisting in recording, sound synthesis and other fields of music production for quite some time. The actual performance of music continues to be an area in which human players are chosen over computer performers. Musical performance is an area in which personalization is more important than consistency. Human players play with each other, reacting to phrases and ideas created by the players that they are playing with. Computer performers lack the ability to react to the changes in the performance that humans perceive naturally, giving the human players an advantage over the computer performers. This thesis creates a framework for describing unique musical performers that can play along in realtime with human players. FrTime, a reactive programming language, is used to constantly create new musical phrases. Musical phrases are constructed by unique user programmed performers and by chord changes that the framework provides. The reactive language creates multiple musical phrases for each point in time. A simple module which chooses musical phrases to be performed at the time of performance is created.

This thesis introduces a new methodology for robust form parameter design of complex hull form geometry via constraint programming, automatic differentiation, interval arithmetic, and truncated hierarchical B- splines. To date, there has been no clearly stated methodology for assuring consistency of general (equality and inequality) constraints across an entire geometric form parameter ship hull design space. In contrast, the method to be given here can be used to produce guaranteed narrowing of the design space, such that infeasible portions are eliminated. Furthermore, we can guarantee that any set of form parameters generated by our method will be self consistent. It is for this reason that we use the title feasible form parameter design. In form parameter design, a design space is represented by a tuple of design parameters which are extended in each design space dimension. In this representation, a single feasible design is a consistent set of real valued parameters, one for every component of the design space tuple. Using the methodology to be given here, we pick out designs which consist of consistent parameters, narrowed to any desired precision up to that of the machine, even for equality constraints. Furthermore, the method is developed to enable the generation of complex hull forms using an extension of the basic rules idea to allow for automated generation of rules networks, plus the use of the truncated hierarchical B-splines, a wavelet-adaptive extension of standard B-splines and hierarchical B-splines. The adaptive resolution methods are employed in order to allow an automated program the freedom to generate complex B-spline representations of the geometry in a robust manner across multiple levels of detail. Thus two complementary objectives are pursued: ensuring feasible starting sets of form parameters, and enabling the generation of complex hull form geometry.

In the field of speech recognition, an algorithm must learn to tell the difference between "a nice rock" and "a gneiss rock". These identical-sounding phrases are called oronyms. Word frequency dictionaries are often used by speech recognition systems to help resolve phonetic sequences with more than one possible orthographic phrase interpretation, by looking up which oronym of the root phonetic sequence contains the most-common words.Our paper demonstrates a technique used to validate word frequency dictionary values. We chose to use frequency values from the UNISYN dictionary, which tallies each word on a per-occurance basis, using a proprietary text corpus, to calculate word frequency.In the first phase of our user study, we generated oronym strings for the phrase "a nice cold hour", and had over a dozen people make 62 of the most-common oronyms for that phrase. In the second phase, we selected 15 of the phase one recordings, and had 74 different people transcribe each one, for a total of 953 transcriptions overall. If the frequency dictionary values for our test phrases accurately reflected the real-world expectations of actual listeners, we would expect that the most-commonly transcribed phrases in our user study would roughly correspond with our metric for the most likely oronym interpretation of the root phrase. During the course of our study, we found that using per-occurance frequency values, like those found in the UNISYN dictionary, when computing our overall-phrase-frequency metric caused the end result to be thrown off by excessively common words, such as "the", "is", and "a" These super-common words had such high per-occurance tallies that they overpowered any effect that any regular word had on a frequency metric. When we used frequency values from the COCA dictionary, which has word frequency values tallied on a document-count basis instead of a UNISYN-like per-occurance basis, we found that this effect was mitigated. As a result, we do not recommend using the UNISYN dictionary for word frequency purposes.

With the shift to learn and consume information through our mobile devices, most academic research is still only presented in long-form text. The Stanford Scholar Initiative has explored the segment of content creation and consumption of academic research through video. However, there has been another popular shift in presenting information from various social media platforms and media outlets in the past few years. Snapchat and Instagram have introduced the concept of tappable “Stories” that have gained popularity in the realm of content consumption. To accelerate the growth of the creation of these research talks, I propose an alternative to video: a tappable Snapchat-like interface. This style is achieved using AMP, Google’s open source project to optimize web experiences on mobile, and particularly the AMP Stories visual medium. My research seeks to explore how the process and quality of consuming the content of academic papers would change if instead of watching videos, users would consume content through Stories on mobile instead. Since this form of content consumption is still largely unresearched in the academic context, I approached this research with a human-centered design process, going through a few iterations to test various prototypes before formulating research questions and designing an experiment. I tested various formats of research consumption through Stories with pilot users, and learned many lessons to iterate from along the way. I created a way to consume research papers in a Stories format, and designed a comparative study to measure the effectiveness of consuming research papers through the Stories medium and the video medium. The results indicate that Stories are a quicker way to consume the same content, and improve the user’s pace of comprehension. Further, the Stories medium provides the user a self-paced method—both temporally and content-wise—to consume technical research topics, and is deemed as a less boring method to do so in comparison to video. While Stories gave the learner a chance to actively participate in consumption by tapping, the video experience is enjoyed because of its reduced effort and addition of an audio component. These findings suggest that the Stories medium may be a promising interface in educational contexts, for distributing scientific content and assisting with active learning.

There is a growing societal need to address the increasing prevalence of behavioral health issues, such as obesity, alcohol or drug use, and general lack of treatment adherence for a variety of health problems. The statistics, worldwide and in the USA, are daunting. Excessive alcohol use is the third leading preventable cause of death in the United States (with 79,000 deaths annually), and is responsible for a wide range of health and social problems. On the positive side though, these behavioral health issues (and associated possible diseases) can often be prevented with relatively simple lifestyle changes, such as losing weight with a diet and/or physical exercise, or learning how to reduce alcohol consumption. Medicine has therefore started to move toward finding ways of preventively promoting wellness, rather than solely treating already established illness. Evidence-based patient-centered Brief Motivational Interviewing (BMI) interven- tions have been found particularly effective in helping people find intrinsic motivation to change problem behaviors after short counseling sessions, and to maintain healthy lifestyles over the long-term. Lack of locally available personnel well-trained in BMI, however, often limits access to successful interventions for people in need. To fill this accessibility gap, Computer-Based Interventions (CBIs) have started to emerge. Success of the CBIs, however, critically relies on insuring engagement and retention of CBI users so that they remain motivated to use these systems and come back to use them over the long term as necessary. Because of their text-only interfaces, current CBIs can therefore only express limited empathy and rapport, which are the most important factors of health interventions. Fortunately, in the last decade, computer science research has progressed in the design of simulated human characters with anthropomorphic communicative abilities. Virtual characters interact using humans’ innate communication modalities, such as facial expressions, body language, speech, and natural language understanding. By advancing research in Artificial Intelligence (AI), we can improve the ability of artificial agents to help us solve CBI problems. To facilitate successful communication and social interaction between artificial agents and human partners, it is essential that aspects of human social behavior, especially empathy and rapport, be considered when designing human-computer interfaces. Hence, the goal of the present dissertation is to provide a computational model of rapport to enhance an artificial agent’s social behavior, and to provide an experimental tool for the psychological theories shaping the model. Parts of this thesis were already published in [LYL+12, AYL12, AL13, ALYR13, LAYR13, YALR13, ALY14].

Dans le domaine des environnements numériques, un enjeu particulier consiste à construire des systèmes permettant aux utilisateurs de partager et de réutiliser leurs expériences. Cette thèse s'intéresse à la problématique générale des recommandations contextuelles pour des applications web dans un contexte particulier : tâche complexes, beaucoup de données, différents types d'utilisateurs (du débutant au professionnel), etc. Nous cherchons à fournir une assistance à l'utilisateur en prenant en compte le contexte et la dynamique des tâches que l'utilisateur effectue. On cherche à fournir des recommandations dynamiques qui sont enrichies au fur et à mesure des expériences. Pour fournir ces recommandations dynamiques, nous nous appuyons sur le Raisonnement à Partir de l'Expérience Tracée (RàPET). Dans le RàPET, les traces d'interaction constituent d'importants conteneurs de connaissances. Ces traces permettent de mieux comprendre le comportement des utilisateurs et leurs activités. Par conséquent, elles représentent également le contexte de l'activité. Les traces peuvent donc venir nourrir un assistant à partir d'expérience en lui fournissant des connaissances appropriées. Dans cette thèse, nous présentons un état de l'art sur les systèmes d'assistances dynamiques et nous rappelons les concepts généraux des systèmes à base de traces. Nous avons proposé une formalisation des traces modélisées et des processus qui permettent de manipuler ces traces. Nous avons notamment défini une méthode pour établir des mesures de similarité afin de comparer des traces modélisées. Nous avons implémenté ces propositions dans un outil appelé TStore. Cet outil permet le stockage, la transformation, la gestion et la réutilisation des traces modélisées. Ensuite, nous avons proposé un mécanisme de rejouage de traces pour permettre aux utilisateurs de revenir à un état précédent de l'application. Enfin, nous avons décrit une approche de recommandations à partir de traces. Le moteur de recommandations est alimenté par les traces d'interactions laissée par les précédents utilisateurs de l'application. Cette approche s'appuie sur les mesures de similarité proposées plus haut. Nous avons validé nos contributions théoriques à l'aide de deux applications web : SAP BusinessObjects Explorer pour l'analyse de données, et Wanaclip pour la génération semi-automatique de clips vidéos. Le mécanisme de rejouage de traces est démontré dans SAP BusinessObjects Explorer. Les recommandations à base de traces sont illustrées dans l'application Wanaclip. Dans la dernière partie du manuscrit, nous mesurons les performances de TStore et la qualité des recommandations et des mesures de similarité qu'il implémente. Nous discutons aussi des résultats du sondage que nous avons appliqué aux utilisateurs de Wanaclip pour mesurer leur satisfaction. Nos évaluations montrent que notre approche offre des recommandations satisfaisantes et un bon temps de réponse<br>In the field of digital environments, a particular challenge is to build systems that enable users to share and reuse their experiences. In this thesis, we are interested in the general problem of contextual recommendations for specific web applications in a particular context: complex tasks, huge amount of data, various types of users (from novice to professional), etc. We focus on providing user assistance which takes into account the context and the dynamics of users’ tasks. We seek to provide dynamic recommendations that are enriched by new experiences over time. To provide these dynamic recommendations, we make use of Trace-Based Reasoning (TBR). TBR is a recent artificial intelligence paradigm that draws its inspiration from Case-Based Reasoning. In TBR, interaction traces act as an important knowledge container. They help to understand users’ behaviors and their activities. Therefore, they reflect the context of the activity. Traces can feed an experience-based assistant with the adequate and appropriate knowledge. In this thesis, we introduce a state of the art about dynamic assistance systems and the general concepts of Trace-Based Systems. In order to provide experience-based assistance, we have made several contributions. First, we propose a formal representation of modeled traces and a description of the processes involved in their manipulation. Notably, we define a method for computing similarity measures for comparing modeled traces. These proposals have been implemented in a framework named TStore for the storage, transformation, management, and reuse of modeled traces. Next, we describe a trace replay mechanism enabling users to go back to a particular state of the application. This mechanism supports impact propagation of changes during the replay process. Last, we define a recommendation approach based on interaction traces. The recommendation engine is fed by interaction traces left by previous users of the application and stored in a manager, such as TStore. This approach facilitates knowledge sharing between communities of users and relies, among other things, on the similarity measures mentioned above. We have validated our theoretical contributions on two different web applications: SAP BusinessObjects Explorer for data reporting and Wanaclip for generating video clips. The trace replay mechanism is demonstrated in SAP BusinessObjects. Trace-Based Reasoning recommendations are illustrated with Wanaclip to guide users in both video selection, and the actions to perform in order to make quality video clips. In the last part of this manuscript, we measure the performances of TStore and the quality of recommendations and similarity measures implemented in TStore. We also discuss the results of the survey that the users of Wanaclip answered in order to measure their satisfaction. Our evaluations show that our approach offers satisfactory recommendations and good response time

碩士<br>國立高雄應用科技大學<br>模具工程系碩士班<br>101<br>The main purpose of the study is to develop a human-machine interface for artificial hearts. We established an artificial heart cardiac output monitor and diagnostic unit (COMDU). The signal curves and data can be shown and recorded. On the other hand, the artificial heart driver is controlled by solenoid valve and directional valve, to produce heart beat along with significant systolic ratio. The human-machine interface is easy to operate, compile, manage and record data. An animal experiment has been participated to verify that human-machine interface of the COMDU for artificial hearts can clearly provide measurement message for doctors to perform further diagnosis during surgery.

As robots become more prevalent in our everyday lives, both in our workplaces and in our homes, it becomes increasingly likely that people who are not experts in robotics will be asked to interface with robotic devices. It is therefore important to develop robotic controls that are intuitive and easy for novices to use. Robotic hands, in particular, are very useful, but their high dimensionality makes creating intuitive human-machine interfaces for them complex. In this dissertation, we study the control of non-anthropomorphic robotic hands by non-roboticists in two contexts: collaborative manipulation and assistive robotics. In the field of collaborative manipulation, the human and the robot work side by side as independent agents. Teleoperation allows the human to assist the robot when autonomous grasping is not able to deal sufficiently well with corner cases or cannot operate fast enough. Using the teleoperator’s hand as an input device can provide an intuitive control method, but finding a mapping between a human hand and a non-anthropomorphic robot hand can be difficult, due to the hands’ dissimilar kinematics. In this dissertation, we seek to create a mapping between the human hand and a fully actuated, non-anthropomorphic robot hand that is intuitive enough to enable effective real-time teleoperation, even for novice users. We propose a low-dimensional and continuous teleoperation subspace which can be used as an intermediary for mapping between different hand pose spaces. We first propose the general concept of the subspace, its properties and the variables needed to map from the human hand to a robot hand. We then propose three ways to populate the teleoperation subspace mapping. Two of our mappings use a dataglove to harvest information about the user's hand. We define the mapping between joint space and teleoperation subspace with an empirical definition, which requires a person to define hand motions in an intuitive, hand-specific way, and with an algorithmic definition, which is kinematically independent, and uses objects to define the subspace. Our third mapping for the teleoperation subspace uses forearm electromyography (EMG) as a control input. Assistive orthotics is another area of robotics where human-machine interfaces are critical, since, in this field, the robot is attached to the hand of the human user. In this case, the goal is for the robot to assist the human with movements they would not otherwise be able to achieve. Orthotics can improve the quality of life of people who do not have full use of their hands. Human-machine interfaces for assistive hand orthotics that use EMG signals from the affected forearm as input are intuitive and repeated use can strengthen the muscles of the user's affected arm. In this dissertation, we seek to create an EMG based control for an orthotic device used by people who have had a stroke. We would like our control to enable functional motions when used in conjunction with a orthosis and to be robust to changes in the input signal. We propose a control for a wearable hand orthosis which uses an easy to don, commodity forearm EMG band. We develop an supervised algorithm to detect a user’s intent to open and close their hand, and pair this algorithm with a training protocol which makes our intent detection robust to changes in the input signal. We show that this algorithm, when used in conjunction with an orthosis over several weeks, can improve distal function in users. Additionally, we propose two semi-supervised intent detection algorithms designed to keep our control robust to changes in the input data while reducing the length and frequency of our training protocol.

Mestrado de dupla diplomação com a UTFPR - Universidade Tecnológica Federal do Paraná<br>A comunicação é considerada uma ferramenta de importante uso em todos os tipos de relações, que pode ser realizada de forma verbal ou não verbal. Tratando-se de uma pessoa surda, a linguagem gestual portuguesa (LGP - Portugal), ou linguagem brasileira de sinais (Libras - Brasil), contribui para a comunicação dos surdos, que enfrentam barreiras que dificultam sua comunicação, em locais de trabalho, vida social entre outros. Considerando a grande comunidade de surdos no mundo, que necessitam de ajuda para se comunicar e que tenham facilidade em realizar essa comunicação, e também, tendo em conta o avanço da tecnologia nos últimos anos, surgiu a possibilidade da criação de interfaces humanocomputador, mais simples e de fácil acesso e compreensão. Neste contexto, e com o uso da Inteligência Artificial, o principal objetivo deste trabalho é estudar a possibilidade de criar uma ferramenta que possa realizar a identificação e tradução das letras do alfabeto para computador através da criação de um sistema composto pela captura e tradução de gestos a partir de um Kinect mais o Software MatLab. Com os estudos realizados neste trabalho, observou-se que o dispositivo possui precisão de 96.71% em relação ao gestos treinados e apresenta a possibilidade de realizar testes e tradução em tempo real.<br>Communication is considered a tool of important use in all types of relationships, which can be carried out verbally or non-verbally. In the case of a deaf person, Portuguese sign language (LGP - Portugal), or Brazilian sign language (Libras - Brazil), contributes to their communication, who face barriers communication, in workplaces, life among others. Considering the large community of deaf people in the world who needs help to communicate and easily communicates, and also, taking into account the advancement of technology in recent years, the possibility of creating human-computer interfaces came up, simpler and easy to access and understand. In this context, and with the use of Artificial Intelligence, the main objective of this work was to study the possibility of creating a tool that can perform the identification and translation of the letters of the alphabet to a computer through the creation of a system composed by the capture and translation of gestures from a Kinect plus MatLab Software. With the studies carried out in this work, it was observed that the device has 96.71% accuracy in relation to the trained gestures and the possibility to carry out tests and real time translation.

Ce mémoire concerne la télé-échographie robotisée, vue comme un ensemble de plusieurs modules (poste maître, poste esclave, communication, image, visio-conférence, retour haptique, etc.). Le travail s'inscrit dans le cadre de la conception, du développement et de l'évaluation d.un tel système et concerne plus spécialement le poste esclave. Une présentation générale de la télé-échographie robotisée est fournie afin d.ensuite se focaliser sur l'architecture du robot esclave dont deux prototypes TER1 et TER2 sont présentés. Le robot porte-sonde possède 6ddl (degrés de liberté). Le premier prototype peut être considéré comme un robot parallèle ayant une structure pour la translation du robot et une autre pour son orientation. Le deuxième prototype est considéré comme un robot hybride dont la structure pour la translation est un sous-système parallèle et celle de l.orientation peut être considérée comme une structure série. Les deux prototypes sont des robots non rigides avec transmission par câbles. Le coeur de ce travail concerne cette nouvelle architecture qui est proposée tout en respectant les consignes de légèreté, compliance et portabilité. En effet, un des objectifs de ce système est de pouvoir être utilisé là où d'habitude il n'existe pas d'experts médicaux qui puissent réaliser l'examen échographique. Un modèle géométrique inverse est développé pour chaque prototype afin de déduire les longueurs des câbles dont on peut déduire les déplacements (translation et orientation) du robot sur le corps du patient. Pour le premier prototype, une loi de commande du type PIDA (Proportionnelle, Intégrale et Dérivative avec un terme d'Anticipation) est utilisée. Le robot TER2 utilise une commande articulaire en boucle fermée. Des tests et des expériences in-vitro et in-vivo ont été réalisés et sont présentés dans ce mémoire, incluant des expériences locales et à distance (Grenoble-Toulouse, Grenoble-Brest, Grenoble-Toulon). Des résultats encourageants et l'avis favorable des médecins pour l'utilisation de ce type de système lors des expériences ont été obtenus