Integración de tecnologías no invasivas para promover la autonomía y mejorar la calidad de vida de personas frágiles
Fecha
2024-04-08
Autores
Título de la revista
ISSN de la revista
Título del volumen
Editor
Jaén : Universidad de Jaén
Resumen
Esta tesis doctoral nace para mejorar la autonomía y calidad de vida de personas frágiles mediante un sistema de tecnologías no invasivas y personalizadas, atendiendo a requisitos como baja invasividad, accesibilidad económica, autonomía, escalabilidad, integración ambiental, y fácil despliegue, con un énfasis en la privacidad y seguridad del usuario.
Propone un sistema distribuido que incorpora sensores multimodales, como dispositivos de audio, visión, localización, y sensores ambientales, junto con modelos avanzados de reconocimiento de actividades adaptados tanto a individuos como a espacios con múltiples ocupantes. Se emplea Lógica Difusa para modelos basados en conocimiento y Aprendizaje Profundo para modelos orientados a datos, facilitando la adaptación a distintas situaciones cotidianas en entornos inteligentes.
Cada una de las propuestas de sensorización y trazabilidad de usuarios se ha evaluado en casos de estudio que permiten explorar su viabilidad, sensibilidad y precisión en entornos reales.
This doctoral thesis aims to improve the autonomy and quality of life of frail individuals through a system of non-invasive and personalized technologies, addressing requirements such as low invasiveness, economic accessibility, autonomy, scalability, environmental integration, and easy deployment, with an emphasis on user privacy and security. It proposes a distributed system that incorporates multimodal sensors, such as audio, vision, location, and environmental sensors, along with advanced activity recognition models adapted for both individuals and spaces with multiple occupants. Fuzzy Logic is used for knowledge-based models and Deep Learning for data-oriented models, facilitating adaptation to various daily situations in smart environments. Each of the sensorization and user traceability proposals has been evaluated in case studies that explore their viability, sensitivity, and accuracy in real -world environments.
This doctoral thesis aims to improve the autonomy and quality of life of frail individuals through a system of non-invasive and personalized technologies, addressing requirements such as low invasiveness, economic accessibility, autonomy, scalability, environmental integration, and easy deployment, with an emphasis on user privacy and security. It proposes a distributed system that incorporates multimodal sensors, such as audio, vision, location, and environmental sensors, along with advanced activity recognition models adapted for both individuals and spaces with multiple occupants. Fuzzy Logic is used for knowledge-based models and Deep Learning for data-oriented models, facilitating adaptation to various daily situations in smart environments. Each of the sensorization and user traceability proposals has been evaluated in case studies that explore their viability, sensitivity, and accuracy in real -world environments.
Descripción
Palabras clave
Reconocimiento de actividades, personas frágiles, multiocupación, entornos inteligentes, sensores de baja invasividad.
Citación
p.[http://hdl.handle.net/10953/]