Color constancy is the ability of the human visual system to recognize object colors regardless of the scene illumination. The aim of computational color constancy is to make images look as if they were taken under white light. The hardest and most important step here is illumination estimation and for the sake of simplicity, uniform illumination is often assumed.

Many benchmark datasets for uniform illumination estimation have been published. Despite the significant progress that they allowed, each of them had some kind of issues, e.g. non-linear images, images with pronounced non-uniform illumination, several sensors per dataset, arbitrary usage of images, etc. Additionally, for some reported results it is not always fully clear what kind of data splitting produced them.

For this reason, this challenge proposes a reboot in experimental results where a new image dataset is given.

Researchers in the exciting area of illumination estimation are invited to participate in the challenge. For more information please visit the Instructions page.

The best research teams will be required to present their results in the Int'l Workshop on Color Vision that will be organized within the program of the 11th Int'l Symposium on Image and Signal Processing and Analysis.

THE REWARDS for the best solutions that will be presented at the workshop are as follows:

• • 1st place - 1000 EUR

  • 2nd place - 500 EUR

  • 3rd place - 250 EUR