Welcome to the InfoSci Platform
Reference Hub8
Adversarial Reconstruction CNN for Illumination-Robust Frontal Face Image Recovery and Recognition

Adversarial Reconstruction CNN for Illumination-Robust Frontal Face Image Recovery and Recognition

Liping Yang, Bin Yang, Xiaohua Gu
Copyright: © 2021 |Volume: 15 |Issue: 2 |Pages: 16
ISSN: 1557-3958|EISSN: 1557-3966|EISBN13: 9781799859833|DOI: 10.4018/IJCINI.20210401.oa2
Cite Article Cite Article

MLA

Yang, Liping, et al. "Adversarial Reconstruction CNN for Illumination-Robust Frontal Face Image Recovery and Recognition." IJCINI vol.15, no.2 2021: pp.18-33. http://doi.org/10.4018/IJCINI.20210401.oa2

APA

Yang, L., Yang, B., & Gu, X. (2021). Adversarial Reconstruction CNN for Illumination-Robust Frontal Face Image Recovery and Recognition. International Journal of Cognitive Informatics and Natural Intelligence (IJCINI), 15(2), 18-33. http://doi.org/10.4018/IJCINI.20210401.oa2

Chicago

Yang, Liping, Bin Yang, and Xiaohua Gu. "Adversarial Reconstruction CNN for Illumination-Robust Frontal Face Image Recovery and Recognition," International Journal of Cognitive Informatics and Natural Intelligence (IJCINI) 15, no.2: 18-33. http://doi.org/10.4018/IJCINI.20210401.oa2

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

This article proposes an adversarial reconstruction convolution neural network (ARCNN) for non-uniform illumination frontal face image recovery and recognition. The proposed ARCNN includes a reconstruction network and a discriminative network. The authors employ GAN framework to learn the reconstruction network in an adversarial manner. This article integrates gradient loss and perceptual loss terms, which are able to preserve the detailed and spatial structure image information, into the overall reconstruction loss function to constraint the reconstruction procedure. Experiments are conducted on the typical illumination-sensitive dataset, extended YaleB dataset. The reconstructed results demonstrate that the proposed ARCNN approach can remove the illumination and shadow information and recover natural uniform illuminated face image from non-uniform illuminated ones. Face recognition results on the extended YaleB dataset demonstrate that the proposed ARCNN reconstruction procedure can also preserve the discriminative information of face image for classification task.