Robust eye gaze estimation


by J Wísniewska, M Rezaei, R Klette
Abstract:
Eye gaze detection under challenging lighting conditions is a non-trivial task. Pixel intensity and the shades around the eye region may change depending on the time of day, location, or due to artificial lighting. This paper introduces a lighting-adaptive solution for robust eye gaze detection. First, we propose a binarization and cropping technique to limit our region of interest. Then we develop a gradient-based method for eye-pupil detection; and finally, we introduce an adaptive eye-corner detection technique that altogether lead to robust eye gaze estimation. Experimental results show the outperformance of the proposed method compared with related techniques.
Reference:
Robust eye gaze estimation (J Wísniewska, M Rezaei, R Klette), In Computer Vision and Graphics, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Springer Verlag, volume 8671, 2014.
Bibtex Entry:
@inproceedings{wsniewska2014robustestimation,
author = "Wísniewska, J and Rezaei, M and Klette, R",
booktitle = "Computer Vision and Graphics, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
organization = "Warsaw",
pages = "636--644",
publisher = "Springer Verlag",
title = "Robust eye gaze estimation",
volume = "8671",
year = "2014",
abstract = "Eye gaze detection under challenging lighting conditions is a non-trivial task. Pixel intensity and the shades around the eye region may change depending on the time of day, location, or due to artificial lighting. This paper introduces a lighting-adaptive solution for robust eye gaze detection. First, we propose a binarization and cropping technique to limit our region of interest. Then we develop a gradient-based method for eye-pupil detection; and finally, we introduce an adaptive eye-corner detection technique that altogether lead to robust eye gaze estimation. Experimental results show the outperformance of the proposed method compared with related techniques.",
startyear = "2014",
startmonth = "Sep",
startday = "15",
finishyear = "2014",
finishmonth = "Sep",
finishday = "17",
issn = "0302-9743",
eissn = "1611-3349",
language = "eng",
conference = "International Conference in Computer Vision and Graphics",
}