Please use this identifier to cite or link to this item:
Title: Deep periocular representation aiming video surveillance.
Authors: Moreira, Gladston Juliano Prates
Luz, Eduardo José da Silva
Zanlorensi Junior, Luiz Antonio
Gomes, David Menotti
Keywords: Deep learning
Transfer learning
VGG Periocular region
Video surveillance
Issue Date: 2017
Citation: MOREIRA, G. J. P. et al. Deep periocular representation aiming video surveillance. Pattern Recognition Letters, v. 114, p. 2-12, 2018. Disponível em: <>. Acesso em: 16 jun. 2018.
Abstract: Usually, in the deep learning community, it is claimed that generalized representations that yielding out- standing performance / effectiveness require a huge amount of data for learning, which directly affect biometric applications. However, recent works combining transfer learning from other domains have sur- mounted such data application constraints designing interesting and promising deep learning approaches in diverse scenarios where data is not so abundant. In this direction, a biometric system for the peri- ocular region based on deep learning approach is designed and applied on two non-cooperative ocular databases. Impressive representation discrimination is achieved with transfer learning from the facial do- main (a deep convolutional network, called VGG) and fine tuning in the specific periocular region domain. With this design, our proposal surmounts previous state-of-the-art results on NICE (mean decidability of 3.47 against 2.57) and MobBio (equal error rate of 5.42% against 8.73%) competition databases.
ISSN:  01678655
Appears in Collections:DECOM - Artigos publicados em periódicos

Files in This Item:
File Description SizeFormat 
  Restricted Access
2,44 MBAdobe PDFView/Open

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.