RGB to 3D garment reconstruction using UV map representations

Predicting the geometry of a 3D object from just a single image or viewpoint is an intrinsic human feature extremely challenging for machines. For years, in an attempt to solve this problem, different computer vision approaches and techniques have been investigated. One of the domains in which there...

Descripción completa

Detalles Bibliográficos
Autor: Rial Farràs, Albert
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:2117/356875
Acceso en línea:https://hdl.handle.net/2117/356875
Access Level:acceso abierto
Palabra clave:Computer vision
Deep learning
Artificial intelligence
Reconstrucció 3D
Mapes UV
Aprenentatge profund
Visió per computador
Intel·ligència artificial
3D reconstruction
UV maps
Visió per ordinador
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
id ES_080149e551cdbd71834566f021ceabb7
oai_identifier_str oai:upcommons.upc.edu:2117/356875
network_acronym_str ES
network_name_str España
repository_id_str
spelling RGB to 3D garment reconstruction using UV map representationsRial Farràs, AlbertComputer visionDeep learningArtificial intelligenceReconstrucció 3DMapes UVAprenentatge profundVisió per computadorIntel·ligència artificial3D reconstructionUV mapsDeep learningComputer visionArtificial intelligenceVisió per ordinadorAprenentatge profundIntel·ligència artificialÀrees temàtiques de la UPC::Informàtica::Intel·ligència artificialPredicting the geometry of a 3D object from just a single image or viewpoint is an intrinsic human feature extremely challenging for machines. For years, in an attempt to solve this problem, different computer vision approaches and techniques have been investigated. One of the domains in which there has been more research has been the 3D reconstruction and modelling of human bodies. However, the greatest advances in this field have concentrated on the recovery of unclothed human bodies, ignoring garments. Garments are highly detailed, dynamic objects made up of particles that interact with each other and with other objects, making the task of reconstruction even more difficult. Therefore, having a lightweight 3D representation capable of modelling fine details is of great importance. This thesis presents a deep learning framework based on Generative Adversarial Networks (GANs) to reconstruct 3D garment models from a single RGB image. It has the peculiarity of using UV maps to represent 3D data, a lightweight representation capable of dealing with high-resolution details and wrinkles. With this model and kind of 3D representation, we achieve state-of-the-art results on CLOTH3D dataset, generating good quality and realistic reconstructions regardless of the garment topology, human pose, occlusions and lightning, and thus demonstrating the suitability of UV maps for 3D domains and tasks.Universitat Politècnica de CatalunyaEscalera Guerrero, SergioMadadi, Meysam20212021-06-2820212021-11-22master thesishttp://purl.org/coar/resource_type/c_bdccNAhttp://purl.org/coar/version/c_be7fb7dd8ff6fe43info:eu-repo/semantics/masterThesisapplication/pdfhttps://hdl.handle.net/2117/356875reponame:UPCommons. Portal del coneixement obert de la UPCinstname:Universitat Politècnica de Catalunya (UPC)Inglésengopen accesshttp://purl.org/coar/access_right/c_abf2info:eu-repo/semantics/openAccessoai:upcommons.upc.edu:2117/3568752026-05-27T15:37:01Z
dc.title.none.fl_str_mv RGB to 3D garment reconstruction using UV map representations
title RGB to 3D garment reconstruction using UV map representations
spellingShingle RGB to 3D garment reconstruction using UV map representations
Rial Farràs, Albert
Computer vision
Deep learning
Artificial intelligence
Reconstrucció 3D
Mapes UV
Aprenentatge profund
Visió per computador
Intel·ligència artificial
3D reconstruction
UV maps
Deep learning
Computer vision
Artificial intelligence
Visió per ordinador
Aprenentatge profund
Intel·ligència artificial
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
title_short RGB to 3D garment reconstruction using UV map representations
title_full RGB to 3D garment reconstruction using UV map representations
title_fullStr RGB to 3D garment reconstruction using UV map representations
title_full_unstemmed RGB to 3D garment reconstruction using UV map representations
title_sort RGB to 3D garment reconstruction using UV map representations
dc.creator.none.fl_str_mv Rial Farràs, Albert
author Rial Farràs, Albert
author_facet Rial Farràs, Albert
author_role author
dc.contributor.none.fl_str_mv Escalera Guerrero, Sergio
Madadi, Meysam
dc.subject.none.fl_str_mv Computer vision
Deep learning
Artificial intelligence
Reconstrucció 3D
Mapes UV
Aprenentatge profund
Visió per computador
Intel·ligència artificial
3D reconstruction
UV maps
Deep learning
Computer vision
Artificial intelligence
Visió per ordinador
Aprenentatge profund
Intel·ligència artificial
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
topic Computer vision
Deep learning
Artificial intelligence
Reconstrucció 3D
Mapes UV
Aprenentatge profund
Visió per computador
Intel·ligència artificial
3D reconstruction
UV maps
Deep learning
Computer vision
Artificial intelligence
Visió per ordinador
Aprenentatge profund
Intel·ligència artificial
Àrees temàtiques de la UPC::Informàtica::Intel·ligència artificial
description Predicting the geometry of a 3D object from just a single image or viewpoint is an intrinsic human feature extremely challenging for machines. For years, in an attempt to solve this problem, different computer vision approaches and techniques have been investigated. One of the domains in which there has been more research has been the 3D reconstruction and modelling of human bodies. However, the greatest advances in this field have concentrated on the recovery of unclothed human bodies, ignoring garments. Garments are highly detailed, dynamic objects made up of particles that interact with each other and with other objects, making the task of reconstruction even more difficult. Therefore, having a lightweight 3D representation capable of modelling fine details is of great importance. This thesis presents a deep learning framework based on Generative Adversarial Networks (GANs) to reconstruct 3D garment models from a single RGB image. It has the peculiarity of using UV maps to represent 3D data, a lightweight representation capable of dealing with high-resolution details and wrinkles. With this model and kind of 3D representation, we achieve state-of-the-art results on CLOTH3D dataset, generating good quality and realistic reconstructions regardless of the garment topology, human pose, occlusions and lightning, and thus demonstrating the suitability of UV maps for 3D domains and tasks.
publishDate 2021
dc.date.none.fl_str_mv 2021
2021-06-28
2021
2021-11-22
dc.type.none.fl_str_mv master thesis
http://purl.org/coar/resource_type/c_bdcc
NA
http://purl.org/coar/version/c_be7fb7dd8ff6fe43
dc.type.openaire.fl_str_mv info:eu-repo/semantics/masterThesis
format masterThesis
dc.identifier.none.fl_str_mv https://hdl.handle.net/2117/356875
url https://hdl.handle.net/2117/356875
dc.language.none.fl_str_mv Inglés
eng
language_invalid_str_mv Inglés
language eng
dc.rights.none.fl_str_mv open access
http://purl.org/coar/access_right/c_abf2
dc.rights.openaire.fl_str_mv info:eu-repo/semantics/openAccess
rights_invalid_str_mv open access
http://purl.org/coar/access_right/c_abf2
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Universitat Politècnica de Catalunya
publisher.none.fl_str_mv Universitat Politècnica de Catalunya
dc.source.none.fl_str_mv reponame:UPCommons. Portal del coneixement obert de la UPC
instname:Universitat Politècnica de Catalunya (UPC)
instname_str Universitat Politècnica de Catalunya (UPC)
reponame_str UPCommons. Portal del coneixement obert de la UPC
collection UPCommons. Portal del coneixement obert de la UPC
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869403017232515072
score 15.198674

Similares en LA Referencia