Gaze-based Action Zones: A universal interaction modality for IoT devices

In order to employ gaze interaction with real-world appliances, techniques to trigger actions and manipulate object states are required. In this work we adapt two previously introduced ideas of one-stroke gaze gestures to gain a universal interaction modality usable for distributed IoT de- vices. We therefore investigate two different types of action zones, that are located (i) at the borders of an object or (ii) at the borders of the users’ field-of-view. Once an object is selected by using dwell time selection, action zones are activated and object state manipulation can be performed by gazing into such an area. A pilot user study was executed to compare the two different placements of action zones and while it highlights that both types lead to highly accurate and promising results, in specific settings one type may be preferable to the other.