Implementation of our paper: Sensor-Independent Illumination Estimation for DNN Models, BMVC, 2019.
Learning a new canonical space in an unsupervised manner allows us to train a single deep model on multiple camera sensors in order to estimate scene illuminant captured by a new unseen sensor.
Mahmoud Afifi and Michael S. Brown. Sensor-Independent Illumination Estimation for DNN Models. In British Machine Vision Conference (BMVC), 2019
support Matlab 2019b or higher
paper info is added