Video Task Effect Eye-Tracking
The eye-tracking data are publicly available to the research community. Please cite the following references if you use this database in your research.
- H. Alers, H. Liu, J. Redi and I. Heynderickx, "TUD Video Quality Database: Eye-Tracking Release 2", http://mmi.tudelft.nl/iqlab/video_task_eye_tracking_1
- H. Alers, H. Liu, J. Redi and I. Heynderickx, "Examining the effect of task on viewing behavior in videos using saliency maps", IS&T/SPIE Electronic Imaging 2012, Human Vision and Electronic Imaging XVII, Jan 2012.
Videos + MOS quality scores: TUD_Video_Task_EyeTracking.zip
Maps: Maps Group1 (scoring) [2.2 GB]
Maps: Maps Group2 (scoring) [1.8 GB]
Maps: Maps Group3 (Free-looking) [2.0 GB]
Maps: Maps Group4 (Free-looking) [4.1 GB]
The first file is password protected. To get the password you can contact Hani Alers (firstname.lastname@example.org)
Please note that the file-sizes are quite big and may require long download times!
Eye-tracking data were collected in order to better understand how people watch videos when assessing their quality level. In the experiment, two groups of viewers were shown video segments and were given two different tasks. One was to simply look at the video as if they were watching TV or an internet downloaded video. The other was to assess the quality level of the videos. With the eye tracker, it was possible to track the eye movement of the viewers and map the salient regions for images given the different tasks and with different levels of video quality.
More details on the experimental set-up and results can be found in the references given above.