Development and Evaluation of a Multi-Functional Gaze-Controlled Interface for Physically Challenged Computer Users

When we interact with computers today, we typically feed input into the computer with our hands controlling a mouse or keyboard, while we employ our eyes to examine the computer’s output on a screen. Intuitively, it seems beneficial to this process to also use the eyes as input devices- by measuring eye movements indicating shifts of attention – and thus eliminate the motor control loop for arm and hand. Obviously, gaze-controlled interfaces are particularly useful for physically challenged users, especially those who cannot move their limbs and also have difficulties using speech-recognition systems. The aim of the present study is the development and empirical evaluation of such a gaze-controlled interface that allows their physically challenged users to easily perform everyday tasks such as writing text, writing and reading e-mails, and doing calculations. Obviously, using the eyes as a means of input differs from using a standard input device – such as a computer mouse – in many ways. On the one hand, eye movements can serve as a fast and direct means of controlling a computer, but on the other hand, they are not completely under conscious control, so when using gaze-control interfaces we may sometimes trigger actions unintentionally (the “Midas-Touch Problem”). It is a crucial point in the design of gaze-controlled interfaces to minimize such undesirable effects. Consequently, the present study includes not only the development of the interface, but also its empirical evaluation. In cooperation with the Center for Adaptive Computing at UMass Boston, physically challenged users will test the interface and give valuable feedback for the improvement of our system. Several different kinds of gaze-controlled interfaces have already been developed for physicall