Learning to Predict Where People Look

Tilke Judd, Krista Ehinger, Fr├ędo Durand, Antonio Torralba

Here we show the eye tracking data we collected from 15 different users on 1003 different images. This ground truth data was then used to learn a model of where people look (see our paper). Click on a thumbnail to see the larger image. Press play to view the scanpath from each of our viewers.

This is raw data.
Red vs Yellow dots.

Viewer 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15