%0 Journal Article
%T Deep intrinsic decomposition trained on surreal scenes yet with realistic light effects
%A Hassan Ahmed Sial
%A Ramon Baldrich
%A Maria Vanrell
%J Journal of the Optical Society of America A
%D 2020
%V 37
%N 1
%F Hassan Ahmed Sial2020
%O CIC; 600.140; 600.12; 600.118
%O exported from refbase (http://refbase.cvc.uab.es/show.php?record=3311), last updated on Tue, 09 Feb 2021 10:58:53 +0100
%X Estimation of intrinsic images still remains a challenging task due to weaknesses of ground-truth datasets, which either are too small or present non-realistic issues. On the other hand, end-to-end deep learning architectures start to achieve interesting results that we believe could be improved if important physical hints were not ignored. In this work, we present a twofold framework: (a) a flexible generation of images overcoming some classical dataset problems such as larger size jointly with coherent lighting appearance; and (b) a flexible architecture tying physical properties through intrinsic losses. Our proposal is versatile, presents low computation time, and achieves state-of-the-art results.
%U https://doi.org/10.1364/JOSAA.37.000001
%U http://refbase.cvc.uab.es/files/SBV2019.pdf
%P 1-15