Deep intrinsic decomposition trained on surreal scenes yet with realistic light effects
This work solves the problem of realistic intrinsic decomposition for computer vision researchers, though it appears incremental by combining dataset generation with physical losses.
The paper tackles the challenge of intrinsic image decomposition by addressing weaknesses in ground-truth datasets and ignoring physical hints in deep learning methods, achieving state-of-the-art results with low computation time.
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.