Enhancing Attention through the Eye Tracking

The contribution describes an implementation of a system capable of measuring and possibly improving the human attention through the use of eye-tracking techniques. The used techniques achieve high accuracy for non-head mounted cameras with low resolution in visible spectrum. In order to achieve this, an analysis of existing methods was carried out. The implementation of the system is created through the utilization of Python-compatible library for face detection as well as using Timm and Barth algorithm for pupil detection. The main focus of the testing phase is the accuracy of the face detection and pupil detection algorithms in various lighting conditions as well as under different view angles. Additional tests are performed on participants with different eye-colors. The experimental solution may be used also for tracking of the attention of students or in various educational games.