Title
Finding an Efficient Threshold for Fixation Detection in Eye Gaze Tracking.
Abstract
We propose a combined analytical/statistical method to determine an efficient threshold on the dispersion of estimates of the point of gaze POG to indicate a user fixation. The experimental data for this study was obtained with an EyeTech TM3 eye gaze tracker EGT. The experimental protocol to make the user fixate on pre-determined visual targets was implemented using the C language and OpenCV. Subjects first used the system in a training mode, from which an individualized dispersion threshold was obtained. Our approach was verified by applying the individualized threshold to POG data from a second run, in testing mode, with encouraging results.
Year
DOI
Venue
2016
10.1007/978-3-319-39516-6_9
HCI
Field
DocType
Citations 
Point of gaze,Computer vision,Experimental data,Computer science,Eye tracking,Artificial intelligence
Conference
0
PageRank 
References 
Authors
0.34
4
5