This data is available as a supplement to the following paper. If you use this data, please cite:
Zoya Bylinskii, Phillip Isola, Constance Bainbridge, Antonio Torralba, Aude Oliva. "Intrinsic and Extrinsic Effects on Image Memorability", Vision Research (2015)
Project page: http://figrim.mit.edu/index_eyetracking.html
Provided are 630 target images and over 2K filler images, all with eye fixations. All images have been sampled from 21 different indoor and outdoor scene categories.
Target images consist of 30 images per scene category. Filler images consist of 94-105 images per scene category.
Eye fixations are included for an average of 16 observers (min: 11, max: 22) per target image and an average of 15 observers (min: 11, max: 19) per filler images.
Target images have been seen by each participant 3 times and the fixations on all 3 image repetitions are included. Filler images have only been seen by each participant once.
The target images are additionally provided along with full LabelMe object annotations.