Increasing performance of strabismus detection models through the synthesis of photorealistic images of the eye region Aumento da performance de modelos de detecção de estrabismo através da síntese de imagens fotorrealistas da região dos olhos

Página do item simplificado
dc.contributor.authorSantos J.S.
dc.contributor.authorFrango I.S.
dc.date.accessioned2024-03-12T23:47:10Z
dc.date.available2024-03-12T23:47:10Z
dc.date.issued2020
dc.description.abstract© 2020, Associacao Iberica de Sistemas e Tecnologias de Informacao. All rights reserved.This study, using a Deep Convolutional Generative Adversarial Networks technique, develops a generative model capable of generating photorealistic images of the eye region that can be used as objects of a data augmentation approach to increase the accuracy of strabismus detection models. This paper contains a simple experiment, a strabismus detector model, which shows a 7% accuracy gain when using synthetic images as a data augmentation.
dc.description.firstpage93
dc.description.issuenumberE33
dc.description.lastpage104
dc.description.volume2020
dc.identifier.issn1646-9895
dc.identifier.urihttps://dspace.mackenzie.br/handle/10899/34935
dc.relation.ispartofRISTI - Revista Iberica de Sistemas e Tecnologias de Informacao
dc.rightsAcesso Restrito
dc.subject.otherlanguageGAN; DCGAN
dc.subject.otherlanguageImage synthesis
dc.subject.otherlanguageStrabismus
dc.titleIncreasing performance of strabismus detection models through the synthesis of photorealistic images of the eye region Aumento da performance de modelos de detecção de estrabismo através da síntese de imagens fotorrealistas da região dos olhos
dc.typeArtigo
local.scopus.citations0
local.scopus.eid2-s2.0-85094594995
local.scopus.updated2024-05-01
local.scopus.urlhttps://www.scopus.com/inward/record.uri?partnerID=HzOxMe3b&scp=85094594995&origin=inward
Arquivos
Coleções
Artigos de periódico