Spelling suggestions: "subject:"high technology."" "subject:"igh technology.""
211 |
Para além do movimento maker: um contraste de diferentes tendências em espaços de construção digital na Educação / Beyond maker movement: a contrast of different trends in digital construction spaces in EducationSilva, Rodrigo Barbosa e 31 August 2017 (has links)
Com o recente crescimento do interesse em atividades manuais baseada em tecnologias digitais, espaços de construção começam a fazer parte de projetos educacionais públicos e privados no país. Estes espaços são planejados, construídos e executados seguindo tendências transnacionais de construção digital. Esta tese contrasta propostas maker baseadas em FabLabs, na Maker Media, em críticas à tecnoutopia californiana e no FabLearn com objetivo de compreender as origens, objetivos e implicações de cada uma dessas diferentes propostas de tecnologias para a Educação. Considerando a sub-teorização do movimento maker em geral, que privilegia o fazer acima do refletir, esta tese apresenta conceitos de tecnologia baseado em Álvaro Vieira Pinto, de práxis e liberdade em Paulo Freire, e de bases social da técnica no campo de Ciência, Tecnologia e Sociedade. Trata-se de uma pesquisa descritiva de fatos e fenômenos em voga na tecnociência e educação brasileiras. Os resultados alcançados são uma abordagem crítico-reflexiva das diferentes vertentes maker, o contraste entre diferentes propostas de construção digital, uma contribuição a propostas progressistas de educação e a valorização e expansão da obra de pensadores nacionais de Educação e tecnologias. Conclui-se que a proposta FabLearn é condicente com ideias freirianas para Educação e que parte da falta de embasamento teórico do movimento maker em geral pode ser preenchida pelo pensamento filosófico de Álvaro Vieira Pinto e educacional de Paulo Freire, em uma perspectiva emancipatória e inclusiva da sociedade. / Spaces dedicated to construction based on digital resources are taking part of public and private educational projects in Brazil as a result of the growing public interest in hands-on activities. These spaces are planned, built, and executed under transnational trends of digital construction. This Ph.D. thesis contrasts maker proposals, e.g. FabLabs, Maker Media, criticism to the Californian tecnoutopia, and FabLearn, in order to comprehend the origins, the aims and the implications of each one of these distinct proposals of technologies in Education. One considering the under-theorization of maker movement in general, which privileges the ´making´ over reflection, this thesis presents concepts of technology based on Alvaro Vieira Pinto, and praxis and freedom on Paulo Freire’s work, along a discussion of social basis of techniques on Science, Technology and Society Studies. It is a descriptive research about facts and phenomenons in an ongoing debate about Brazilian techno science and education. As results, it presents firstly a critical reflexive approach of diverse maker proposals, secondly a contribution to progressive education discussions, and thirdly the enrichment and expansion of national thinkers’ theories on technology and education. It concludes that FabLearn is consistent with Freire’s ideas of progressive education, and Alvaro Vieira Pinto’s philosophical and Paulo Freire’s educational thoughts can be filled in, based on an emancipatory and inclusive perspective of society, the theoretical gap of maker movement.
|
212 |
Revision of an artificial neural network enabling industrial sortingMalmgren, Henrik January 2019 (has links)
Convolutional artificial neural networks can be applied for image-based object classification to inform automated actions, such as handling of objects on a production line. The present thesis describes theoretical background for creating a classifier and explores the effects of introducing a set of relatively recent techniques to an existing ensemble of classifiers in use for an industrial sorting system.The findings indicate that it's important to use spatial variety dropout regularization for high resolution image inputs, and use an optimizer configuration with good convergence properties. The findings also demonstrate examples of ensemble classifiers being effectively consolidated into unified models using the distillation technique. An analogue arrangement with optimization against multiple output targets, incorporating additional information, showed accuracy gains comparable to ensembling. For use of the classifier on test data with statistics different than those of the dataset, results indicate that augmentation of the input data during classifier creation helps performance, but would, in the current case, likely need to be guided by information about the distribution shift to have sufficiently positive impact to enable a practical application. I suggest, for future development, updated architectures, automated hyperparameter search and leveraging the bountiful unlabeled data potentially available from production lines.
|
Page generated in 0.0701 seconds