For many applications in graphics, design, and human computer interaction, it is essential to understand where humans look in a scene. Where eye tracking devices are not a viable option, models of saliency can be used to predict fixation locations. Most saliency approaches are based on bottom-up computation that does not consider top-down image semantics and often does not match actual eye movements. The database contains collected eye tracking data of 15 viewers on 1003 images and use this database as training and testing examples to learn a model of saliency based on low, middle and high-level image features. This large database of eye tracking data is publicly available with this paper.
ZIP archives with eye tracking database: Set of stimuli: Link: http://people.csail.mit.edu/tjudd/WherePeopleLook/ALLSTIMULI.zip Eye tracking data: http://people.csail.mit.edu/tjudd/WherePeopleLook/DATA.zip Human fixation maps: http://people.csail.mit.edu/tjudd/WherePeopleLook/ALLFIXATIONMAPS.zip
References and Citation
TED09: Tilke, J., Ehinger, K., Durand, F., Torralba, A. Learning to predict where humans look, 2009, Proceedings of the IEEE International Conference on Computer Vision , art. no. 5459462, pp. 2106-2113.