Project: Eye tracking system with a video camera
An eye tracker is a device capable of determining the direction of
gaze of a human observer. Current eye tracking technology is cumbersome
to the user, prohibitively expensive for home or business use,
requires extensive training and/or does not allow head movements. We
present a system that only requires a video camera and computer
software; unlike all other video-based systems we are aware of, it
does not require infra-red or other light-emitting devices. Our
algorithm first locates a few key features of the user's face (pupils
and eye corners) using orientation correlation, and then feeds that
location data to a 2-layer neural network with just a few (10) hidden
units which, after training it with only 85 images, is able to locate
the point on a screen that the user is looking at. A preliminary
version of our system performs at present with under .15 units mean
error, where units are 3 cm apart from each other. The user's head is
free to move as long as both eyes remain in focus and in view of the
camera.
- Hardware: CCD Black and white camera, no zoom, Frame grabber,Pentium
- Software: C, C++, Windows programming.
Bibliography
-
Colin Ware and Harutune H. Mikaelian, "An Evaluation of an Eye Tracker
as a Device for Computer Input", Proceedings, SIGCHI+GI'87 Conference, Toronto,1987
-
A. J. Glenstrup and T. Engell-Nielsen, "Eye Controlled Media: Present and Future State",http://www.diku.dk/\%7Epanic/eyegaze/article.html,1995
-
DeDean A. Pomerleau, " Neural Network Perception for Mobil Robot Guidance", Kluwer Academic Publishers, 1993
-
John G. Daugman "High confidence visual recognition of persons
by a test of statistical independence ", IEEE Trans. Patt. Anal. &
Mach. Intellig. 1993, Vol 15, No 11,1148-1161
-
B. Simard, B. Prasada and M.K. Sinha, "On-line character recognition
using handwriting modelling", pr, 1993, VOL 26, 993-1007
-
Robert J. K. Jacob,"The Use of Eye Movements in Human-Computer
Interaction Techniques: What You Look At is What You Get", ACM
Transactions on Information Systems,1991, VOL 9,152-169
-
Baluja, S. and Pomerleau, D.A.},"Non-Intrusive Gaze Tracking Using
Artificial Neural Networks ",Advances in Neural Information
Processing Systems, San Mateo, CA: Morgan Kaufmann Publishers ,1994,VOL 6,753-760