Using artificial intelligence for predictive eye-tracking analysis to evaluate photographs

Abstract

The goal of the study was to determine how close the eye-tracking results predicted by the AI model are to actual measurements and whether they can be used in scientific research or in real business cases. The study was based on a carefully prepared photo database of 30 photos of varying complexity and colour. The photos were shown to 110 participants (age and gender evenly distributed), and eye-tracking device (Tobii X120) was used to measure how the photos were viewed. In comparison, the same photos were tested using an AI-based application (Expoze.io). The final results show the comparison between the heatmaps and transparent gaze visualisations of the collected data with the two used measurement methods. Suggestions are made in which cases and how the two described methods should be used.