The ability to learn is essential for robots to function and perform services within a dynamic human environment. Robot programming by demonstration facilitates learning through a human teacher without the need to develop new code for each task that the robot performs. In order for learning to be generalizable, the robot needs to be able to grasp the underlying structure of the task being learned. This requires appropriate knowledge abstraction and representation. The goal of this thesis is to develop a learning by imitation system that abstracts knowledge of human demonstrations of a task and represents the abstracted knowledge in a hierarchical framework. The learning by imitation system is capable of performing both action and object recognition based on video stream data at the lower level of the hierarchy, while the sequence of actions and object states observed is reconstructed at the higher level of the hierarchy in order to form a coherent representation of the task. Furthermore, error recovery capabilities are included in the learning by imitation system to improve robustness to unexpected situations during task execution. The first part of the thesis focuses on motion learning to allow the robot to both recognize the actions for task representation at the higher level of the hierarchy and to perform the actions to imitate the task. In order to efficiently learn actions, the actions are segmented into meaningful atomic units called motion primitives. These motion primitives are then modeled using dynamic movement primitives (DMPs), a dynamical system model that can robustly generate motion trajectories to arbitrary goal positions while maintaining the overall shape of the demonstrated motion trajectory. The DMPs also contain weight parameters that are reflective of the shape of the motion trajectory. These weight parameters are clustered using affinity propagation (AP), an efficient exemplar clustering algorithm, in order to determine groups of similar motion primitives and thus, performing motion recognition. The approach of DMPs combined with APs was experimentally verified on two separate motion data sets for its ability to recognize and generate motion primitives. The second part of the thesis outlines how the task representation is created and used for imitating observed tasks. This includes object and object state recognition using simple computer vision techniques as well as the automatic construction of a Petri net (PN) model to describe an observed task. Tasks are composed of a sequence of actions that have specific pre-conditions, i.e. object states required before the action can be performed, and post-conditions, i.e. object states that result from the action. The PNs inherently encode pre-conditions and post-conditions of a particular event, i.e. action, and can model tasks as a coherent sequence of actions and object states. In addition, PNs are very flexible in modeling a variety of tasks including tasks that involve both sequential and parallel components. The automatic PN creation process has been tested on both a sequential two block stacking task and a three block stacking task involving both sequential and parallel components. The PN provides a meaningful representation of the observed tasks that can be used by a robot to imitate the tasks. Lastly, error recovery capabilities are added to the learning by imitation system in order to allow the robot to readjust the sequence of actions needed during task execution. The error recovery component is able to deal with two types of errors: unexpected, but known situations and unexpected, unknown situations. In the case of unexpected, but known situations, the learning system is able to search through the PN to identify the known situation and the actions needed to complete the task. This ability is useful not only for error recovery from known situations, but also for human robot collaboration, where the human unexpectedly helps to complete part of the task. In the case of situations that are both unexpected and unknown, the robot will prompt the human demonstrator to teach how to recover from the error to a known state. By observing the error recovery procedure and automatically extending the PN with the error recovery information, the situation encountered becomes part of the known situations and the robot is able to autonomously recover from the error in the future. This error recovery approach was tested successfully on errors encountered during the three block stacking task.

This thesis contains four essays that analyze the behaviors that evolve when populations of boundedly rational individuals interact strategically for a long period of time. Individuals are boundedly rational in the sense that their strategy choices are determined by simple rules of adaptation -- learning rules. Convergence results for general finite games are first obtained in a homogenous setting, where all populations consist either of stochastic imitators, who almost always imitate the most successful strategy in a sample from their own population's past strategy choices, or stochastic better repliers, who almost always play a strategy that gives at least as high expected payoff as a sample distribution of all populations' past play. Similar results are then obtained in a heterogeneous setting, where both of these learning rules are represented in each population. It is found that only strategies in certain sets are played in the limit, as time goes to infinity and the mutation rate tends to zero. Sufficient conditions for the selection of a Pareto efficient such set are also provided. Finally, the analysis is extended to natural selection among learning rules. The question is whether there exists a learning rule that is evolutionarily stable, in the sense that a population employing this learning rule cannot be invaded by individuals using a different rule. Monte Carlo simulations for a large class of learning rules and four different games indicate that only a learning rule that takes full account of hypothetical payoffs to strategies that are not played is evolutionarily stable in almost all cases. / Diss. Stockholm : Handelshögsk., 2001

Thesis (Ph. D.)--University of Oregon, 2007. / Typescript. Includes vita and abstract. Includes bibliographical references (leaves 276-292). Also available for download via the World Wide Web; free to University of Oregon users.

De senaste åren har små barns användning av digitala medier ökat och det finns idag en uppsjö av medieinnehåll riktat till denna målgrupp. Forskning visar dock att små barn har svårt att generalisera lärande från skärmar till verkligheten. Samtittande (co-viewing) med vuxna och verbala ledtrådar har föreslagits kunna underlätta lärandet men forskningen är ännu begränsad. Föreliggande studie syftade därför till att undersöka huruvida instruerad co-viewing har någon inverkan på deklarativt minne, mätt med fördröjd imitation, hos barn vid två års ålder. För att undersöka detta användes experimentell mellangruppsdesign, där deltagarna (N = 35) tillsammans med sina föräldrar fick se en film på surfplatta där uppgifter ur ett deklarativt minnestest (FIT-24) presenterades. Deltagarna randomiserades till experimentgrupp, där föräldrarna instruerades att beskriva händelseförloppet i filmen för barnen, och kontrollgrupp, där föräldrarna instruerades att titta på filmen tillsammans med barnen såsom de skulle gjort hemma. Efter en fördröjning testades vad barnen lärt sig av minnesuppgifterna. Utöver minnestestet genomfördes screening av utvecklingsnivå (Bayley-S).Resultatet visade ingen signifikant skillnad i deklarativt minne mellan betingelserna. Däremot återfanns positiva signifikanta samband mellan deklarativt minne och generell utveckling samt språklig förmåga. Därtill klarade deltagare i båda grupperna att generalisera kunskap från surfplatta till verklighet, vilket visar att tvååriga barn kan lära sig från digitala medier.

Il est possible de mener une étude sur les « Parterres » considérés comme des « lieux communs » de la fabrique des Jardins de plaine de la Renaissance, au sens rhétorique, fédérateur et prospectif du terme. Comme les tropes d'un discours, ils sont répartis sur la « plate forme » du jardin afin d'agir sur les émotions, dans les catégories rhétoriques et poétiques de l'amplificatio, l'inventio, la dispositio, l'elocutio, l'imitatio et la copia. Ces ornements sont des moyens d'expression au service d'un vaste projet d'harmonisation décorative de la demeure aristocratique et de son environnement à toutes les échelles matérielles et imbriquées ; cette coordination formelle doit satisfaire la double exigence de concinnitas et de varietas comme principes de plaisir et de bienséance. Afin d'expliciter la conception des jardins intégrés à ce « grand dessein », les artistes font intervenir une représentation cognitive par analogie (paradeigma) impliquant le revêtement décoratif des sols, soit la ressemblance de la composition du jardin avec celle d'un « tapis de Turquie » orné de « Moresques ». La diffusion de ces formes est assurée par des suites de modèles imprimées dont nous faisons la bibliographie critique. Nous abordons aussi leur mise en œuvre au jardin qui est obtenue par la compartimentation de la matière première végétale déclinée dans les catégories changeantes et mélangées du « Verger », de « Préau », du « Potager », du « Jardin de Simples », et du « Bouquetier » / It is possible to undertake a study on the "Paterres" (Flower beds) regarded as 'common places" of the invention of the plain Gardens of the Renaissance, in the rhetorical, inclusive and forward sense of term. Like the Tropes of a speech, they are divided on the "plate forme" of the garden to act on emotions in poetic and rhetorical categories of amplificatio, inventio, dispositio, elocutio, imitatio and copia. These ornaments are a means of expression in the service of a larger project of decorative harmonization of a place and its environment at all scales and imbricated material, which must satisfy the dual requirement of concinnitas and varietas as principles of pleasure ad propriety. In order to clarify the design of gardens included in this "grand design", artists involve a cognitive representation by analogy (paradeigma) involving decorative coating of soil or the resemblance of the composition of the garden with a "Turkey carpet" decorated with "Moresques". The diffusion of these forms is provided by sequences of models of which we create a critical bibliography. Their implementation in the garden is achieved by compartmentalization of vegetable raw materials declined in changing and mixed categories of "Orchad", "Medow", "Potager", "Jardin de simples" and "Bouquetier"

Em interações humano-robô ainda existem muitas limitações a serem superadas referentes à provisão de uma comunicação natural quanto aos sentidos humanos. A capacidade de interagir com os seres humanos de maneira natural em contextos sociais (pelo uso da fala, gestos, expressões faciais, movimentos do corpo) é um ponto fundamental para garantir a aceitação de robôs em uma sociedade de pessoas não especialistas em manipulação de engenhos robóticos. Outrossim, a maioria dos robôs existentes possui habilidades limitadas de percepção, cognição e comportamento em comparação com seres humanos. Nesse contexto, este projeto de pesquisa investigou o potencial da arquitetura robótica do humanoide NAO, no tocante à capacidade de realizar interações com seres humanos através de imitação de movimentos do corpo de uma pessoa e pelo controle do robô. Quanto a sensores, foi utilizado um sensor câmera não-intrusivo de profundidade incorporado no dispositivo Kinect. Quanto às técnicas, alguns conceitos matemáticos foram abordados para abstração das configurações espaciais de algumas junções/membros do corpo humano essas configurações foram capturadas por meio da utilização da biblioteca OpenNI. Os experimentos realizados versaram sobre a imitação e o controle do robô por meio da avaliação de vários usuários. Os resultados desses experimentos revelaram um desempenho satisfatório quanto ao sistema desenvolvido / In human-robot interactions there are still many restrictions to overcome regarding the provision of a communication as natural to the human senses. The ability to interact with humans in a natural way in social contexts (the use of speech, gestures, facial expressions, body movements) is a key point to ensure the acceptance of robots in a society of people not specialized in manipulation of robotic devices. Moreover, most existing robots have limited abilities of perception, cognition and behavior in comparison with humans. In this context, this research project investigated the potential of the robotic architecture of the NAO humanoid robot, in terms of ability to perform interactions with humans through imitation of body movements of a person and the robot control. As for sensors, we used a non-intrusive sensor depth-camera built into the device Kinect. As to techniques, some mathematical concepts were discussed for abstraction of the spatial configurations of some joints/members of the human body these configurations were captured through the use of the OpenNI library. The performed experiments were about imitation and the control of the robot through the evaluation of various users. The results of these experiments showed a satisfactory performance for the developed system

Made available in DSpace on 2016-03-15T19:40:20Z (GMT). No. of bitstreams: 1 Ana Paula Melillo Lourencao.pdf: 1005274 bytes, checksum: 5e3229d34c2a7d215e3c3d36f613b46e (MD5) Previous issue date: 2015-02-06 / Imitation is an important function for child development. From there the child connects with the outside world to observe the other in their attitudes, words and actions. Studies show the link between imitation and social interaction, language acquisition, motor and cognitive skills. Children with Autism Spectrum Disorder (ASD) have difficulties in social interaction and to connect with the outside world. The general objective of this study was to compare the performance in imitation of tasks between children with Autism Spectrum Disorder and children with typically development. The instruments used to compare the performances were: Imitation battery (BI) with 13 tasks developed for children between 12 and 24 months and the other battery d130 (Bd130) with 15 tasks developed for children between 2 and 5 years, these instruments have been applied in 3 groups of children: A: typical children aged between 18 and 24 months -BI; B: typical children 6 years old - Bd130 and C: children with Autism Spectrum Disorders (ASD) aged between 3 and 7 years-BI and Bd130. These batteries have been applied by the researcher individually and in a quiet environment. The result showed that: 1) the two sets of tasks could be applied and are reliable to evaluate the ability of imitation of children. The d130 battery should better assess typical children younger than 6 years because almost all managed to perform all tasks, with little difficulty in some more complex items; 2) the tasks of BI for typical children between 18 and 24 months are well structured and performance is a direct function of age. Increase the sample size aged at least every two months will be needed to better demonstrate the ability to imitate by age and by gender. The same tasks for children with ASD showed capacity of discrimination, indicating the commitment of this sample. These results should be confirmed in a larger number of children; 3) we found that in children between 18 and 24 months the success of the children answer the tasks presented depend on the appropriate link establishment with the applicator; 4) the tasks proposed by d130 battery were suitable to assess the ability of children with ASD imitation between the ages of 3 to 7 years. It was observed that the skills of the act of imitating that more are affected are those related to motor skill as: copy a circle, square, cross and star. It was possible to see a big difficulty in holding the pencil properly to execute their proposed models. Another skill that proved impaired was the imitation of building a simple visual pattern, showing once again that children with ASD have difficulty to pay attention to details. And the last skill that is damaged is the one linked to the language because many of the children did not speak or had difficulty in speech, not being able to recognize the sound they were producing. With increased sample size may be tested the performance of these tasks in a broader manifestations of ASD spectrum; 5) it is proposed that the d130 function - imitation of the ICF-CY (International Functionality Coding for Children and Adolescents) is classified with the use of these 15 tasks d130 Battery / A imitação é uma função relevante para o desenvolvimento infantil. A partir dela a criança se conecta com o mundo exterior ao observar o outro em suas atitudes, falas e ações. Estudos evidenciam a ligação entre imitação e interação social, aquisição de linguagem, habilidades motoras e cognitivas. Crianças com Transtorno do Espectro do Autismo (TEA) apresentam dificuldade na interação social e de se conectar com o mundo exterior. O objetivo deste trabalho foi comparar o desempenho em tarefas de imitação entre crianças com Transtorno do Espectro Autista e crianças com desenvolvimento. Os instrumentos utilizados para comparar os desempenhos foram: Bateria de Imitação (BI) com 13 tarefas desenvolvido para crianças entre 12 e 24 meses e o outro Bateria d130 (Bd130) com 15 tarefas desenvolvido para crianças entre 2 e 5 anos, esses instrumentos foram aplicados a 3 grupos de crianças: A: crianças típicas com idades entre 18 e 24 meses BI; B:crianças típicas com 6 anos de idade Bd130 e C: crianças com Transtornos do Espectro do Autismo (TEA) com idades entre 3 e 7 anos- BI e Bd130. Essas baterias foram aplicadas pelo pesquisador individualmente e em um ambiente tranquilo. Os resultaram permitiram concluir que: 1) os dois conjuntos de tarefas puderam ser aplicados e são confiáveis para avaliar a capacidade de imitação das crianças. A Bateria d130 deve avaliar melhor crianças típicas com idades inferiores a 6 anos pois praticamente todas conseguiram executar todas as tarefas, apresentando pequena dificuldade em alguns itens mais complexos; 2) as tarefas da BI para crianças típicas entre 18 e 24 meses estão bem estruturadas e o desempenho é função direta da idade. Aumento do número amostral com idades variando no máximo a cada 2 meses será necessário para evidenciar melhor a capacidade de imitar por faixa etária e por gênero. As mesmas tarefas para crianças com TEA mostraram capacidade de discriminação, indicando o comprometimento desta amostra. Tais resultados deverão ser confirmados em um numero maior de crianças; 3) foi possível verificar que nas crianças entre 18 e 24 meses o sucesso na resposta das crianças as tarefas apresentadas dependem do estabelecimento de vínculo adequado com o aplicador; 4) as tarefas propostas pela Bateria d130 foram adequadas para avaliar a capacidade de imitação de crianças com TEA entre as idades de 3 a 7 anos. Observou-se que as habilidades do ato de imitar que mais estão prejudicadas são aquelas ligadas à habilidade motora como: copiar um círculo, quadrado, cruz e asterisco. Foi possível perceber uma grande dificuldade em segurar o lápis de maneira correta para execução dos modelos propostos. Outra habilidade que se mostrou prejudicada foi a imitação da construção de um padrão visual simples, mostrando mais uma vez que as crianças com TEA tem dificuldade de se atentar aos detalhes. E a última habilidade que se encontra prejudicada é aquela ligada à linguagem, pois muitas das crianças não falavam ou apresentavam dificuldade na fala, não sendo possível reconhecer o som que elas estavam produzindo. Com aumento do número amostral poderá ser testado o desempenho destas tarefas num espectro de manifestações mais amplo dos TEA; 5) propõe-se que a função d130 imitação da CIF-CJ (Codificação Internacional de Funcionalidade para Crianças e Adolescentes) seja classificada com o uso destas 15 tarefas da Bateria d130.

A presente pesquisa se concentra na recuperação da acepção setecentista dos conceitos de estilo e de gosto na música. Com base em fontes originais do séc. XVIII, mostra como a idéia de gosto veio a se associar à de estilo no decorrer do século XVIII, passando a constituir a base de termos como gostos reunidos e estilo galante, comumente relacionados à música de Georg Philipp Telemann (1681-1767). Este estudo aborda os estilos e gostos adotados por Telemann a partir de uma perspectiva histórica e, para tanto, utiliza as retóricas clássicas, nas quais se apoiaram os autores setecentistas; os tratados sobre gosto dos séc. XVII e XVIII; as poéticas musicais setecentistas; e as autobiografias de Telemann (de 1718 e 1740). Deste modo, propomos o estudo de ambos os conceitos não apenas a partir de sua perspectiva técnica ou nacional, mas recuperando sua função dentro do contexto da elocutio retórica, envolvendo a idéia de decoro. Nas fontes setecentistas, o uso destes termos se relaciona com os sistemas de adequação entre as diferentes maneiras de escrita em relação à ocasião e ao gosto nacional, ambos regulados pelo decoro. Desta forma, estes conceitos constituem aspectos de uma concepção musical única, que pode ser melhor compreendida a partir da perspectiva dupla do gosto e do estilo. No séc. XVIII, o gosto regulava as maneiras de compor, formando estilos próprios de um determinado grupo, país, ou auctoritas. Os dois principais parâmetros de gosto normativos do período são a música francesa e a italiana. Deste modo, a partir da imitação e do estudo dos modelos estrangeiros formou-se a prática denominada gostos reunidos - mencionada nos escritos setecentistas alemães como gemischten Goût, vermischter Geschmack e Gosto Alemão. Deste processo de imitação, onde tomar algo como modelo não se refere a uma cópia do discurso, mas sim à aprendizagem e exercitação de como tratar tecnicamente um material de origens diversas a fim de gerar algo novo, surge um novo gênero, denominado por Telemann e seus contemporâneos como Concert en Ouverture, ou Concertouverture. / This research is focused on recovering the seventeenth-century sense of the concepts style and taste in music. Based on original works from the eighteenth century, presents the process of association between the ideas of taste and style during the eighteenth century, becoming the basis of terms as mixed tastes and gallant style, often used on descriptions of Georg Philipp Telemann (1681-1767) music. This study analyses the styles and tastes used by Telemann from a historical approach, using the classical rethoric in which the seventeenth-century authors based their writings; treatises about taste from the seventeenth and eighteenth century period; musical poetics from the seventeenth century and Telemann´s autobiographies (from 1718 an 1740). This work proposed the study of both concepts not only from the technical or national perspectives, but recovering its place on the rhetoric context of elocutio tied to the notion of decorum. The use of the previous terms in the works from the seventeenth century relates them to the suitability between different ways of composition for each circumstance and national taste, both ruled by decorum. Hence, these concepts are part of an unique musical notion, that can be better understood from the parallel perspectives of taste and style. In eighteenth century, taste ruled composition forms, generating peculiar styles from certain groups, countries and auctoritas. The two main normative parameters of taste in the period were French and Italian music. Thus, from the study and imitation of foreign styles, the practice known as mixed tastes was shaped - it was mentioned in the seventeenth-century German writings as gemischten Goût, vermischter Geschmack and German taste. From this imitation process, where following something as example was not understood as copying the content, but learning and practising the process of working on materials from various sources with the purpose of building something new, a new genre arises. It\'s called by Telemann and his contemporaries Concert en Ouverture, or Concertouverture.

The aim of the thesis was to better understand how second language learners becomes lingual participants in a Swedish preschool. Using the following research questions: In what situations are SLLs lingual participants at the preschool? What does their emerging language look like? Are the participation any different considering children's time in the new L2 context? What kind of support does SLLs receive at the preschool? Two focal children at the same preschool was studied with ethnography of communication as method. The focal children chosen were one girl (Mandarin L1, 4.0 yrs) who arrived to Sweden and started the preschool two months ago at the time of the study, the other a boy (Serbian L1, 3.9 yrs) who started preschool and learning the L2 one year and two months before the study. Situations where the focal children interacted with other children or adults were analyzed.  Results showed that the focal children in interaction with each other understands some of the others intentions before being able to speak about them in the L2. Also episodes of shared attention showed to be the arena for intersubjective lingual negotiation and understanding. Imitation seemed to be the primary strategy for language learning at this stage. These results are in line with Tomasellos (1999, 2003) theory of early language emergence. The focal children became participants in lingual activity when they had the role as Lave and Wengers (1991) legitimate peripheral participants. In interaction with more central participants the focal children imitated their language. However, in interaction with other children the language was centered around formulaic expressions associated with play. These child-child interactions was best described through Rogoffs (2003) notion of guided participation. Whereas the focal children interaction with teachers showed imitation of a more nuanced repertoire of words and was more in line with learning as the scaffolding process put forth by Wood, Bruner & Ross (1976). This leads to a discussion of the connection between language and participation, suggesting a dual model where language development is interconnected with participation in community activities and vice versa. Also the relation between formulaic language, creativity and teacher scaffolding as resources for participation are discussed.

Le but de cette thèse est de proposer des méthodes d'étude et des modèles pour l'analyse des signaux sociaux dans un contexte d'interaction en exploitant à la fois des techniques issues du traitement du signal et de la reconnaissance des formes.Tout d'abord, une méthode non supervisée permettant de mesurer l'imitation entre deux partenaires en termes de délai et de degré est proposée en étudiant uniquement des données gestuelles. Dans un premier temps, des points d'intérêts spatio-temporels sont détectés afin de sélectionner les régions les plus importantes des vidéos. Ils sont ensuite décrits à l'aide d'histogrammes pour permettre la construction de modèles sac-de-mots dans lesquels l'information spatiale est réintroduite. Le degré d'imitation et le délai entre les partenaires sont alors estimés de manière continue grâce à une corrélation-croisée entre les deux modèles sac-de-mots.La deuxième partie de cette thèse porte sur l'extraction automatique d'indices permettant de caractériser des interactions de groupe. Après avoir regroupé tous les indices couramment employés dans la littérature, nous avons proposé l'utilisation d'une factorisation en matrice non négative. En plus d'extraire les indices les plus pertinents, celle-ci a permis de regrouper automatiquement et de manière non supervisée des meetings en 3 classes correspondant aux trois types de leadership tels que définis par les psychologues.Enfin, la dernière partie se focalise sur l'extraction non supervisée d'indices permettant de caractériser des groupes. La pertinence de ces indices, par rapport à des indices ad-hoc provenant de l'état de l'art, est ensuite validée dans une tâche de reconnaissance des rôles. / In a social interaction, we adapt our behavior to our interlocutors. Studying and understanding the underlying mecanisms of this adaptation is the center of Social Signal Processing. The goal of this thesis is to propose methods of study and models for the analysis of social signals in the context of interaction, by exploiting both social processing and pattern recognition techniques. First, an unsupervised method allowing the measurement of imitation between two partners in terms of delay and degree is proposed, only using gestual data. Spatio-temporal interest point are first detected in order to select the most important regions of videos. Then they are described by histograms in order to construct bag-of-words models in which spatial information is reintroduced. Imitation degree and delay between partners are estimated in a continuous way thanks to cross-correlation between the two bag-of-words models. The second part of this thesis focus on the automatic extraction of features permitting to characterizing group interactions. After regrouping all features commonly used in literature, we proposed the utilization of non-negative factorization. More than only extracting the most pertinent features, it also allowed to automatically regroup, and in an unsupervised manner, meetings in three classes corresponding to three types of leadership defined by psychologists. Finally, the last part focus on unsupervised extraction of features permitting to characterize groups. The relevance of these features, compared to ad-hoc features from state of the art, is then validated in a role recognition task.