Gaze control is a project developed for the Innovation Centre for Assistive Technology (IAT Lab) that will enable an electric wheelchair to be driven by eye tracking. This innovative technology is a very interesting advance for people with reduced mobility. While a first prototype using an infrared camera was in the testing phase, it had the disadvantage of not being fully functional outdoors, as the infrared camera was “dazzled” by too much light. To solve this problem, the Icare Institute developed a new prototype in 2019 using computer vision.
Icare’s developments are based on deep learning algorithms analysing the video stream of a standard camera. As a first step, the recognition system focused on locating and tracking the orientation of the nose. This preliminary work proved to be very conclusive and was carried out in view of an Innosuisse research project submission. This next step should allow the development of the interface connecting the electric wheelchair to the eye system and for the Icare Institute to refine its developments to analyse eye tracking.
