Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-Learn
Autor(a) principal: | |
---|---|
Data de Publicação: | 2022 |
Outros Autores: | , , , |
Tipo de documento: | Artigo de conferência |
Idioma: | eng |
Título da fonte: | Repositório Institucional da UNESP |
Texto Completo: | http://dx.doi.org/10.1007/978-3-031-23236-7_51 http://hdl.handle.net/11449/246822 |
Resumo: | Face Recognition is a challenging task present in different applications and systems. An existing challenge is to recognize faces when imaging conditions are adverse, for example when images come from low-quality cameras or when the subject and the camera are far apart, thus impacting the accuracy of these recognizing systems. Super-Resolution techniques can be used to improve both image resolution and quality, hopefully improving the accuracy of the face recognition task. Among these techniques, the actual state-of-the-art uses Generative Adversarial Networks. One promising option is to train Super-Resolution and Face Recognition as one single network, conducting the network to learn super resolution features that will improve its capability when recognizing faces. In the present work, we trained a super resolution face recognition model using a jointly-learn approach, combining a generative network for super resolution and a ResNet50 for Face Recognition. The model was trained with a discriminator network, following the generative adversarial training. The images generated by the network were convincing, but we could not converge the face recognition model. We hope that our contributions could help future works on this topic. Code is publicly available at https://github.com/OliRafa/SRFR-GAN. |
id |
UNSP_2e4fefab7fb6890ea5eba749e01a0469 |
---|---|
oai_identifier_str |
oai:repositorio.unesp.br:11449/246822 |
network_acronym_str |
UNSP |
network_name_str |
Repositório Institucional da UNESP |
repository_id_str |
2946 |
spelling |
Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-LearnFace RecognitionGenerative Adversarial NetworksMachine learningSuper-resolutionFace Recognition is a challenging task present in different applications and systems. An existing challenge is to recognize faces when imaging conditions are adverse, for example when images come from low-quality cameras or when the subject and the camera are far apart, thus impacting the accuracy of these recognizing systems. Super-Resolution techniques can be used to improve both image resolution and quality, hopefully improving the accuracy of the face recognition task. Among these techniques, the actual state-of-the-art uses Generative Adversarial Networks. One promising option is to train Super-Resolution and Face Recognition as one single network, conducting the network to learn super resolution features that will improve its capability when recognizing faces. In the present work, we trained a super resolution face recognition model using a jointly-learn approach, combining a generative network for super resolution and a ResNet50 for Face Recognition. The model was trained with a discriminator network, following the generative adversarial training. The images generated by the network were convincing, but we could not converge the face recognition model. We hope that our contributions could help future works on this topic. Code is publicly available at https://github.com/OliRafa/SRFR-GAN.Instituto Politécnico de BragançaUniversidade Tecnológica Federal do ParanáUniversidade Estadual PaulistaUniversidade Estadual PaulistaInstituto Politécnico de BragançaUniversidade Tecnológica Federal do ParanáUniversidade Estadual Paulista (UNESP)de Oliveira, Rafael AugustoScheeren, Michel HanzenRodrigues, Pedro João SoaresJunior, Arnaldo Candido [UNESP]de Paula Filho, Pedro Luiz2023-07-29T12:51:25Z2023-07-29T12:51:25Z2022-01-01info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/conferenceObject747-762http://dx.doi.org/10.1007/978-3-031-23236-7_51Communications in Computer and Information Science, v. 1754 CCIS, p. 747-762.1865-09371865-0929http://hdl.handle.net/11449/24682210.1007/978-3-031-23236-7_512-s2.0-85147987485Scopusreponame:Repositório Institucional da UNESPinstname:Universidade Estadual Paulista (UNESP)instacron:UNESPengCommunications in Computer and Information Scienceinfo:eu-repo/semantics/openAccess2023-07-29T12:51:25Zoai:repositorio.unesp.br:11449/246822Repositório InstitucionalPUBhttp://repositorio.unesp.br/oai/requestopendoar:29462024-08-05T21:16:27.550113Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP)false |
dc.title.none.fl_str_mv |
Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-Learn |
title |
Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-Learn |
spellingShingle |
Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-Learn de Oliveira, Rafael Augusto Face Recognition Generative Adversarial Networks Machine learning Super-resolution |
title_short |
Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-Learn |
title_full |
Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-Learn |
title_fullStr |
Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-Learn |
title_full_unstemmed |
Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-Learn |
title_sort |
Super-Resolution Face Recognition: An Approach Using Generative Adversarial Networks and Joint-Learn |
author |
de Oliveira, Rafael Augusto |
author_facet |
de Oliveira, Rafael Augusto Scheeren, Michel Hanzen Rodrigues, Pedro João Soares Junior, Arnaldo Candido [UNESP] de Paula Filho, Pedro Luiz |
author_role |
author |
author2 |
Scheeren, Michel Hanzen Rodrigues, Pedro João Soares Junior, Arnaldo Candido [UNESP] de Paula Filho, Pedro Luiz |
author2_role |
author author author author |
dc.contributor.none.fl_str_mv |
Instituto Politécnico de Bragança Universidade Tecnológica Federal do Paraná Universidade Estadual Paulista (UNESP) |
dc.contributor.author.fl_str_mv |
de Oliveira, Rafael Augusto Scheeren, Michel Hanzen Rodrigues, Pedro João Soares Junior, Arnaldo Candido [UNESP] de Paula Filho, Pedro Luiz |
dc.subject.por.fl_str_mv |
Face Recognition Generative Adversarial Networks Machine learning Super-resolution |
topic |
Face Recognition Generative Adversarial Networks Machine learning Super-resolution |
description |
Face Recognition is a challenging task present in different applications and systems. An existing challenge is to recognize faces when imaging conditions are adverse, for example when images come from low-quality cameras or when the subject and the camera are far apart, thus impacting the accuracy of these recognizing systems. Super-Resolution techniques can be used to improve both image resolution and quality, hopefully improving the accuracy of the face recognition task. Among these techniques, the actual state-of-the-art uses Generative Adversarial Networks. One promising option is to train Super-Resolution and Face Recognition as one single network, conducting the network to learn super resolution features that will improve its capability when recognizing faces. In the present work, we trained a super resolution face recognition model using a jointly-learn approach, combining a generative network for super resolution and a ResNet50 for Face Recognition. The model was trained with a discriminator network, following the generative adversarial training. The images generated by the network were convincing, but we could not converge the face recognition model. We hope that our contributions could help future works on this topic. Code is publicly available at https://github.com/OliRafa/SRFR-GAN. |
publishDate |
2022 |
dc.date.none.fl_str_mv |
2022-01-01 2023-07-29T12:51:25Z 2023-07-29T12:51:25Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/conferenceObject |
format |
conferenceObject |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://dx.doi.org/10.1007/978-3-031-23236-7_51 Communications in Computer and Information Science, v. 1754 CCIS, p. 747-762. 1865-0937 1865-0929 http://hdl.handle.net/11449/246822 10.1007/978-3-031-23236-7_51 2-s2.0-85147987485 |
url |
http://dx.doi.org/10.1007/978-3-031-23236-7_51 http://hdl.handle.net/11449/246822 |
identifier_str_mv |
Communications in Computer and Information Science, v. 1754 CCIS, p. 747-762. 1865-0937 1865-0929 10.1007/978-3-031-23236-7_51 2-s2.0-85147987485 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.relation.none.fl_str_mv |
Communications in Computer and Information Science |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
747-762 |
dc.source.none.fl_str_mv |
Scopus reponame:Repositório Institucional da UNESP instname:Universidade Estadual Paulista (UNESP) instacron:UNESP |
instname_str |
Universidade Estadual Paulista (UNESP) |
instacron_str |
UNESP |
institution |
UNESP |
reponame_str |
Repositório Institucional da UNESP |
collection |
Repositório Institucional da UNESP |
repository.name.fl_str_mv |
Repositório Institucional da UNESP - Universidade Estadual Paulista (UNESP) |
repository.mail.fl_str_mv |
|
_version_ |
1808129303332782080 |