Show simple item record

dc.contributorPardàs Feliu, Montse
dc.contributor.authorMartí I Rabadán, Miquel
dc.contributor.otherUniversitat Politècnica de Catalunya. Departament de Teoria del Senyal i Comunicacions
dc.description.abstractMultitask Learning (MTL) was conceived as an approach to improve the generalization ability of machine learning models. When applied to neural networks, multitask models take advantage of sharing resources for reducing the total inference time, memory footprint and model size. We propose MTL as a way to speed up deep learning models for applications in which multiple tasks need to be solved simultaneously, which is particularly useful in embedded, real-time systems such as the ones found in autonomous cars or UAVs. In order to study this approach, we apply MTL to a Computer Vision problem in which both Object Detection and Semantic Segmentation tasks are solved based on the Single Shot Multibox Detector and Fully Convolutional Networks with skip connections respectively, using a ResNet-50 as the base network. We train multitask models for two different datasets, Pascal VOC, which is used to validate the decisions made, and a combination of datasets with aerial view images captured from UAVs. Finally, we analyse the challenges that appear during the process of training multitask networks and try to overcome them. However, these hinder the capacity of our multitask models to reach the performance of the best single-task models trained without the limitations imposed by applying MTL. Nevertheless, multitask networks benefit from sharing resources and are 1.6x faster, lighter and use less memory compared to deploying the single-task models in parallel, which turns essential when running them on a Jetson TX1 SoC as the parallel approach does not fit into memory. We conclude that MTL has the potential to give superior performance as far as the object detection and semantic segmentation tasks are concerned in exchange of a more complex training process that requires overcoming challenges not present in the training of single-task models.
dc.publisherUniversitat Politècnica de Catalunya
dc.rightsS'autoritza la difusió de l'obra mitjançant la llicència Creative Commons o similar 'Reconeixement-NoComercial- SenseObraDerivada'
dc.subjectÀrees temàtiques de la UPC::Enginyeria de la telecomunicació
dc.subject.lcshMachine learning
dc.subject.lcshPattern recognition systems
dc.subject.otherMulti-task Learning
dc.subject.otherDeep Learning
dc.subject.otherComputer vision
dc.subject.otherObject detection
dc.subject.otherSemantic segmentation
dc.titleMulti-task Deep Learning models for real-time deployment in embedded systems
dc.title.alternativeModels multi-tasca basats en xarxes neuronals d’aprenentage profund per al desplegament en sistemes encastats i en temps real
dc.typeMaster thesis
dc.subject.lemacAprenentatge automàtic
dc.subject.lemacReconeixement de formes (Informàtica)
dc.rights.accessOpen Access
dc.audience.mediatorEscola Tècnica Superior d'Enginyeria de Telecomunicació de Barcelona
dc.contributor.covenanteeKungl. tekniska högskolan. Skolan för elektroteknik och datavetenskap

Files in this item


This item appears in the following Collection(s)

Show simple item record

Except where otherwise noted, content on this work is licensed under a Creative Commons license: Attribution-NonCommercial-NoDerivs 3.0 Spain