The DIEM project has produced a series of corpora during its short existence. The parsed eye-movement data and videos are available to download in a format suitable for CARPE from a directory stored on Mediafire. Example CARPE output for each video is available via our Vimeo Channel. There will also be separate postings off this page that will allow an easier way of searching and exploring the available data.
The format of the eye-tracking data is:
[frame] [left_x] [left_y] [left_dil] [left_event] [right_x] [right_y] [right_dil] [right_event]
Frames are 30 frames per second, indexed at 1; x,y are screen coordinates; dil represents pupil dilation; and the event flag represents:
- -1 = Error/dropped frame
- 0 = Blink
- 1 = Fixation
- 2 = Saccade
In visualizing, we only consider the flag 1, when we are sure of fixation.
If you use CARPE in your publication, we would appreciate it if you could cite:
Parag K. Mital, Tim J. Smith, Robin Hill, John M. Henderson. “Clustering of Gaze during Dynamic Scene Viewing is Predicted by Motion” Cognitive Computation, Volume 3, Issue 1, pp 5-24, March 2011.