GAZE-BASED INTERFACES FOR TOUCHLESS HUMAN-MACHINE INTERACTION IN PUBLIC PLACES

The aim of this thesis is the implementation of a calibration-free gaze-based communication for touchless interaction with "devices" in public places. This is particularly relevant for the current COVID-19 emergency, as it allows to interact with computers or machines without having to touch physical screens, buttons, or other input tools, but simply using the gaze as an input channel. As application scenarios, we can think of automatic ticket machines in train stations, information kiosks, scales for weighing fruit and vegetables in supermarkets, etc. The subjects fall into the category of perceptual user interfaces (both for implicit and explicit communication), and in particular in the area of eye tracking interfaces. Eye tracking based interaction systems usually does not work properly without accurate calibration. This thesis focuses on systems that should be easily and quickly used in public contexts, where calibration would be inconvenient. In the developed system, the interaction is performed by looking at specific “buttons”, which provide a visual feedback to the user then perform the associated actions without the need for a physical touch. The developed interfaces have been tested with several participants in different experimental sessions, providing overall satisfying results.

Interfacce basate sullo sguardo per interazione uomo-macchina senza contatto in luoghi pubblici