Search results for " artificial"
showing 10 items of 2657 documents
Contributions and applications around low resource deep learning modeling
2023
El aprendizaje profundo representa la vanguardia del aprendizaje automático en multitud de aplicaciones. Muchas de estas tareas requieren una gran cantidad de recursos computacionales, lo que limita su adopción en dispositivos integrados. El objetivo principal de esta tesis es estudiar métodos y algoritmos que permiten abordar problemas utilizando aprendizaje profundo con bajos recursos computacionales. Este trabajo también tiene como objetivo presentar aplicaciones de aprendizaje profundo en la industria. La primera contribución es una nueva función de activación para redes de aprendizaje profundo: la función de módulo. Los experimentos muestran que la función de activación propuesta logra…
Reinforcement learning in a multi-agent framework for pedestrian simulation
2014
El objetivo de la tesis consiste en la utilización de Aprendizaje por refuerzo (Reinforcement Learning) para generar simulaciones plausibles de peatones en diferentes entornos. Metodología Se ha desarrollado un marco de trabajo multi-agente donde cada agente virtual que aprende un comportamiento de navegación por interacción con el mundo virtual en el que se encuentra junto con el resto de agentes. El mundo virtual es simulado con un motor físico (ODE) que está calibrado con parámetros de peatones humanos extraídos de la bibliografía de la materia. El marco de trabajo es flexible y permite utilizar diferentes algoritmos de aprendizaje (en concreto Q-Learning y Sarsa(lambda) en combinación c…
Análisis de técnicas de “aggregation”/“disaggregation” aplicadas a imágenes satélite para la estimación de parámetros térmicos superficiales a difere…
2023
Las aplicaciones que implican la observación de la superficie terrestre desde plataformas satélites a escala inferior a la regional, como por ejemplo, el caso del seguimiento de cultivos, requieren de una mayor disponibilidad de información térmica, en particular de la temperatura de la superficie terrestre (LST), con resoluciones espaciales apropiadas para un alcance local. Por ello, numerosos autores han propuesto y desarrollado métodos para extraer la LST a nivel “subpíxel”, mediante el empleo de productos complementarios de teledetección, con resultados adecuados para su uso en resoluciones superiores. La mayoría de estos métodos se basan en la correlación entre índices de vegetación, c…
Electrical Safety Analysis in the Presence of Resonant Grounding Neutral
2019
The resonant grounding is one of the possible methods of the system neutral grounding for the medium voltage distribution. IEEE standards do define this grounding configuration, but important advantages and drawbacks of the resonant grounding might not be fully known, due to its rather uncommon application in North America. On the other hand, resonant grounding in Europe is imposed by the increased requirements for power quality, especially for medium-voltage (MV) industrial users, imposed to electric utilities by regulatory authorities, with the purpose to protect the interests of users and consumers. Level of the continuity of the service, magnitude and phase of ground-fault currents, mag…
INTEGRAZIONE, AUTOADATTAMENTO E COSCIENZA ARTIFICIALE
2008
Automated Detection of Microaneurysms Using Scale-Adapted Blob Analysis and Semi-Supervised Learning
2014
International audience; Despite several attempts, automated detection of microaneurysm (MA) from digital fundus images still remains to be an open issue. This is due to the subtle nature of MAs against the surrounding tissues. In this paper, the microaneurysm detection problem is modeled as finding interest regions or blobs from an image and an automatic local-scale selection technique is presented. Several scale-adapted region descriptors are then introduced to characterize these blob regions. A semi-supervised based learning approach, which requires few manually annotated learning examples, is also proposed to train a classifier to detect true MAs. The developed system is built using only…
Image inpainting using directional wavelet packets originating from polynomial splines
2020
The paper presents a new algorithm for the image inpainting problem. The algorithm is using a recently designed versatile library of quasi-analytic complex-valued wavelet packets (qWPs) which originate from polynomial splines of arbitrary orders. Tensor products of 1D qWPs provide a diversity of 2D qWPs oriented in multiple directions. For example, a set of the fourth-level qWPs comprises 62 different directions. The properties of the presented qWPs such as refined frequency resolution, directionality of waveforms with unlimited number of orientations, (anti-)symmetry of waveforms and windowed oscillating structure of waveforms with a variety of frequencies, make them efficient in image pro…
On the use of composite indicators for mobile communications network management in smart sustainable cities
2020
Beyond 5G networks will be fundamental towards enabling sustainable mobile communication networks. One of the most challenging scenarios will be met in ultra-dense networks that are deployed in densely populated areas. In this particular case, mobile network operators should benefit from new assessment metrics and data science tools to ensure an effective management of their networks. In fact, incorporating architectures allowing a cognitive network management framework could simplify processes and enhance the network&rsquo
Computational Rationality as a Theory of Interaction
2022
Funding Information: This work was funded by the Finnish Center for AI and Academy of Finland (“BAD” and “Human Automata”). We thank our reviewers, Xiuli Chen, Joerg Mueller, Christian Guckelsberger, Sebastiaan de Peuter, Samuel Kaski, Pierre-Alexandre Murena, Antti Keuru-lainen, Suyog Chandramouli, and Roderick Murray-Smith for their comments. Publisher Copyright: © 2022 ACM. How do people interact with computers? This fundamental question was asked by Card, Moran, and Newell in 1983 with a proposition to frame it as a question about human cognition - in other words, as a matter of how information is processed in the mind. Recently, the question has been reframed as one of adaptation: how …
Cooperative sound design: a protocol analysis
2016
Formal protocol analysis and linkographic representations are well-established approaches in design cognition studies, in the visual domain. We introduce the method and tools in the auditory domain, by analysing a case of collaborative sound design. We show how they can provide relevant qualitative and quantitative information about the efficiency of the creative process.