Escalera Guerrero, SergioMarín Tur, JavierTylson Baixauli, Emilio2022-05-302022-05-302021-06-30https://hdl.handle.net/2445/186070Treballs finals del Màster de Fonaments de Ciència de Dades, Facultat de matemàtiques, Universitat de Barcelona. Curs: 2020-2021. Tutor: Sergio Escalera Guerrero i Javier Marín Tur[en] This work presents Satellite Style and Structure Generative Adversarial Network (SSGAN), a generative model of high resolution satellite imagery to support image segmentation. Based on spatially adaptive denormalization modules (SPADE) that modulate the activations with respect to segmentation map structure in addition to global descriptor vectors that capture the semantic information in a vector with respect to Open Street Maps (OSM) classes, this model is able to produce consistent aerial imagery. By decoupling the generation of aerial images into a structure map and a carefully defined style vector, we were able to improve the realism and geodiversity of the synthesis with respect to the state-of-the-art baseline. Therefore, the proposed model allows to control the generation not only with respect to the desired structure, but also with respect to a geographic area.39 p.application/pdfengcc-by-nc-nd (c) Emilio Tylson Baixauli, 2021codi: GPL (c) Emilio Tylson Baixauli, 2021http://www.gnu.org/licenses/gpl-3.0.ca.htmlhttp://creativecommons.org/licenses/by-nc-nd/3.0/es/Imatges satel·litàriesVisió per ordinadorAprenentatge automàticTreballs de fi de màsterRemote-sensing imagesComputer visionMachine learningMaster's thesesinfo:eu-repo/semantics/masterThesisinfo:eu-repo/semantics/openAccess