Gaze estimation and analysis of visual scanning patterns with mobile devices

点击翻译按钮获取中文摘要

The goals of this project are to develop methods to measure eye position on a mobile phone and to apply this capability to assess cognitive functions. Currently, systems that measure eye position (eye-trackers) are used in a variety of fields (man-machine interfaces, training, psychology, neurology, etc.) by groups of specialists with access to specialized equipment. By cooperating with a major manufacturer of cell phones, the success of this project will mean that high-quality eye tracking will be available on general-purpose, widely available cell-phone devices. ****To develop high quality eye-tracking systems for cell-phones, one has to overcome significant technical challenges. The first is associated with the development of methods and algorithms to estimate eye-features (e.g., the centre of the pupil) in video images from the cell-phone's front facing camera. Even though we have successfully developed such methods for desk-top systems, the new methods will have to take into account the mechanical, optical and power constraints that are imposed by the cell-phone platforms. The second challenge is that, unlike prior work in which the remote eye-tracker was stationary the eye tracker in this project is free to move relative to the subject. ****Using our desk-top eye-tracking systems we have developed tests to: a) evaluate the effects of injury and the prognosis of recovery from concussion, and b) the probability that mildly cognitively impaired patients will convert to Alzheimer's disease patients. If such tests can be carried out at the site of injury or at the subject's home it will improve the efficacy and benefits of the tests. Successful development of eye-tracking systems for mobile platforms will enable the adaptation and development of a wide range of eye-tracking applications that can significantly benefit the general public.****