Environment perception for micromobility applications

The objective of this project is to see how deep learning technologies, specifically image recognition features and RNNs, can help to make the life of everybody living in the city safer, and help preventing some of the accidents that occur every day, by implementing a functional system capable of de...

Descripción completa

Detalles Bibliográficos
Autor: González López, Julio
Tipo de recurso: tesis de maestría
Fecha de publicación:2021
País:España
Institución:Universitat Politècnica de Catalunya (UPC)
Repositorio:UPCommons. Portal del coneixement obert de la UPC
Idioma:inglés
OAI Identifier:oai:upcommons.upc.edu:2117/360244
Acceso en línea:https://hdl.handle.net/2117/360244
Access Level:acceso abierto
Palabra clave:Deep learning
Neural networks (Computer science)
Vehicles
deep learning
image recognition
micromobility
Aprenentatge profund
Xarxes neuronals (Informàtica)
Àrees temàtiques de la UPC::Enginyeria de la telecomunicació::Telemàtica i xarxes d'ordinadors
Descripción
Sumario:The objective of this project is to see how deep learning technologies, specifically image recognition features and RNNs, can help to make the life of everybody living in the city safer, and help preventing some of the accidents that occur every day, by implementing a functional system capable of detecting the type of lane a Personal Mobility Device (PMD) is circulating in the city of Barcelona. It is also necessary that among the five type of lanes to predict, a higher importance is given to the identification of wether a PMD is circulating or not through the sidewalk, where there is a higher risk of accidents involving pedrestians, and velocity should be reduced more. Besides this, another objective of the thesis will be the implementation, training and evaluation of various deep Recurrent Neural Networks, and posteriorly the comparison with simple convolutional neural networks and viability analysis of these newer technologies applied to PMD.