Concentration detection by eye movements: towards supporting a human

Machines like humanoids or personal computers are needed to assist humans. The tasks of humans become so complicated that machines need to act without detailed orders from humans. Eye movements provide very important information about a human. Many researchers have tried to measure accurate fixation points using an eye tracking system, from which they attempted to extract a human´s orders. With limited resources, it is impossible to calculate accurate fixation points while allowing natural head movement. The authors have attempted to separately analyze the change in face positions and change in direction of gaze without detecting fixation points. This eye tracking system requires 3 features: (1) detection of a face position change and eye movements, (2) wide measurement area to allow for natural head movement, (3) high sampling rate over 20 Hz for detection of saccades. The chosen eye tracking system is based on a stereo camera. To verify the ability to detect whether a human is concentrating or not, experimental measurement is made while a subject is doing several tasks. The results confirm the possibility of detecting a human´s state of attention.