Identity Recognition through Human Gaze Tracking

The paper describes a system for the human machine interaction that is able to identify users according how she looks at the monitor while using a given interface. The system does not need invasive measurements that could limit the naturalness of her actions and detects the eyes movement from the estimation provided by a kinect camera. The proposed approach clusters the sequences of user gaze on the screen characterizing the user identity according the particular pattern his/her gaze follows. The possibility of identify people through gaze movement introduces a new perspective on human-machine interaction. For example, a user can obtain different contents according his recorded preferences and a software can modify its interface to meet the preferences of a given user.