| Sponsor | Irfan Essa
http://www.cc.gatech.edu/~irfan 206 CoC Antonio Haro
|
| Area | CPL / GVU |
Problem
In recent years, researchers and companies have become interested in
perceptual user interfaces; that is, user interfaces that use perception
to support the user interaction. Recently, we created a vision-based system
that detects and tracks eyes robustly in indoor environments. A user interface
that is aware of the number of people interacting with it or in the scene
is valuable in many applications. If your computer knew when you were looking
at it, it would also be able to look at your face and see if you were getting
frustrated, for example.
This project entails creating the foundation for such a system. In it, you will take the existing eye tracker and use the outputs (a list of detected pairs of eyes) to create a user interface mockup on a display tablet that knows when a user is looking at it.
Background
Take a look at an overview of our eye tracker at: http://www.cc.gatech.edu/cpl/projects/pupil/index.html. Also, read our paper on the eye tracker to get an idea of how it works: http://www.cc.gatech.edu/cpl/projects/pupil/pupil.pdf. Note: you do not have to understand the paper fully to do this project; reading it will help you get a high-level understanding of the code, which will be useful since you will be modifying a small portion of it.
What you need to do: