The visual focus of attention dataset is composed of the following parts:

-Videos
	-Baseline conditions (4 experimenters x 2 videos)
	-Adversarial conditions 
		- Natural illumination (4 experimenters x 1 video)
		- Artificial illumination (2 experimenters x 1 video)
		- Dimmed illumination (2 experimenters x 1 video)
		- Motion blur (4 experimenters x 1 video)
- Ground truth
	- Text file containing ground truth
	- Reference images x 2
====================================================================================================
Videos

We collected two sets of videos, comprising 20 sequences, recorded at a rate of 10-fps with 320 x 240 resolution. The whole database constitutes 6691 images in total. 
	
Baseline conditions are used to evaluate the intraclass and interclass generalization capabilities under normal illumination conditions, providing an understanding of the baseline performance, where each of the four experimenters provide two sequences in front of a stationary camera under sufficient natural illumination. 

Adversarial conditions set is composed of 12 video sequences collected from four experimenters under three different conditions. These include artificial illumination, dimmed illumination, and motion blur. For reference purposes, there is one video recorded under natural illumination for each condition. 
====================================================================================================
Ground truth

Ground is organized in the following manner. 
video_name
object_index start_frame end_frame

In other words, the experimenter in video_name.avi focuses on object #object_index between frames #start_frame and #end_frame.

Please check reference images in the same folder to see the correspondence between objects and object indices. 
====================================================================================================

If you use this dataset, please cite the following paper:

Yucel, Z., Salah, A. A., Mericli, C., Mericli, T., Valenti, R., Gevers, T.	 Joint Attention by Gaze Interpolation and Saliency.

@article{yucel2012joint,
  title={Joint Attention by Gaze Interpolation and Saliency},
  author={Y{\"u}cel, Z. and Salah, A.A. and Meri{\c{c}}li, {\c{C}}. and Meri{\c{c}}li, T. and Valenti, R. and Gevers, T.},
  journal  = "IEEE Transactions on Systems, Man, and Cybernetics, Part B: Cybernetics",
  pages    = "1--14",
  year={2012},
  publisher={IEEE}
}