Investigate the challenges of mobile eye-tracking

Abstract

Eye tracking: the process of measuring the point of gaze or the motion of the eye. Behind this definition lies a technology that has been evolving for more than 20 years. Its applications range from marketing and psychology research to substituting the mouse on a computer. Development of the technology has been accompanied by an increase of people using it. Therefore, at the same time that old difficulties where overcome, new challenges were created. Eye tracking can be divided into two groups, depending on how it is used: mobile and non-mobile. A non-mobile setup is a rigid one, with the camera usually placed in a fixed position, in a controlled environment and normally requires calibration before being used. On the other hand, mobile setups are more flexible. For instance, the camera doesn't always have to be in a specific location: it can be sitting on top of the computer screen, or under it, on a support arm, or head-mounted. These kind of gaze trackers are becoming more popular, as they permit more flexibility of use. The consequence of this flexibility is they have to work in environments with changing conditions, so it would be good thing for them not to require calibration every time. This leads to the search of a system, a method that is robust enough to work under varying conditions such as light conditions (number and position of the sources, intensity of the light), indoor or outdoor application, weather conditions (sunny, cloudy). At the same time, that robustness should include the potential users of the system: skin colour, eye colour, morphology of the eye, etc. The aim of this project is to find a solution robust enough to handle the issues mentioned before. Using methods that are already employed, improving them and changing their parameters; taking methods from other fields an applying them to eye tracking; combining both old and new ways of tracking the eye. It could set the foundations for a new line of investigations and development in the area of mobile eye tracking.