May 3-7, 2005
Thesis Presentations
721 Broadway
at Waverly Place
4th Floor
South Elevators
New York, NY 10003
For additional
information email
itp.inquiries@nyu.edu
or call 212.998.1880
| Yarbus 1.0 | ||
|---|---|---|
| Author(s): | Jason Babcock | |
| Instructor: | Barton, Jake | |
| Class: | Final Project Seminar | |
| URL: | http://stage.itp.tsoa.nyu.edu/~jsb314/blog.html | |
| Keywords: | eyetracking, eye movements, generative art | |
| Yarbus 1.0 is a first generation eyetracking system build for OS X Tiger. | ||
| The technology to record eye movements has been around for over a hundred years. Despite rich historical advancements, eye-tracking technology has not surfaced to the mainstream. One of the most fundamental questions in eye movement research is how eyetracking systems should evolve as interfaces between humans and computers. However, that question is difficult to answer because current eyetracking solutions are expensive (~$20K), difficult to use, and rely on poorly designed user interfaces. This thesis examines whether eyetracking really is a plausible interface to computers, and if there is a more affordable, user-friendly, software approach. | ||
| Personal Statement: | The limitations of today’s eye-trackers has become more obvious, and over the last couple of years I have been motivated to start learning how to program with the intention of developing my own custom eye-tracking software. In fact, my purpose for coming to NYU’s Interactive Telecommunications Program was to develop my programming skills in an environment receptive to rapid prototyping. Yarbus 1.0, the software developed for my thesis, is the first eyetracking system build for OS X. | |
| Context: | See Description | |
| Audience: | people interested in eye movements, eye-tracking, or visual perception and mobile computing | |
| User Scenario: | The eyetracking system will be demonstrated with a laptop and will be connected to an external computer via a socket connection. User\'s eye movements will be used to control a generative algorithm on the screen. The software will also be demonstrated in calibration mode where the cursor on the screen indicates where the viewer is looking. | |
| Methodology: | Yarbus 0.1 was developed to handle real-time image capture from two video cameras at 30Hz. Images are rendered to OpenGL textures and various image processing techniques are used to detect the pupil and corneal reflection. The software, developed in C and Objective C (i.e. the Cocoa framework) is designed to run on Apple\'s latest operating system - Tiger. | |
| Sources: | Babcock, J.S. (2005). Yarbus 1.0: Eyetracking Software for OS X. Master’s Thesis. New York: New York University.
Winfield, D., Li, D., Babcock J., Parkhurst D. (2005). Towards an open-hardware open-software toolkit for robust low-cost eye tracking in HCI applications, Iowa State University Human Computer Interaction Technical Report ISU HCI-2005-04 Babcock, J.S., and Pelz, J. (2004). Building a lightweight eyetracking headgear, ETRA 2004: Eye Tracking Research and Applications Symposium. Babcock, J.S., Pelz, J.B., Peak, J.F. (2003). The Wearable Eyetracker: A Tool for the Study of High-level Visual Tasks. Proceedings of the Military Sensing Symposia Specialty Group on Camouflage, Concealment, and Deception, Tucson, Arizona. | |
| Conclusions: | The software developed for this thesis demonstrates that it is possible to do real-time eyetracking with an Apple 12\" laptop and open hardware headgear described in Babcock & Pelz, 2004 (see reference #2). | |