Motivation

This experiment is being carried out as part of my B.Sc. IT (Hons) Artificial Intelligence thesis at the University of Malta. I am being supervised by Dr. Dylan Seychell and co-supervised by Prof. Ing. Carl James Debono.

Saliency is a sub‐problem of computer vision which attempts to tackle the creation of a mathematical model which can approximate visual attention, i.e., the model predicts which objects in an image garner the most prioritisation from the human visual system.

The purpose of this experiment is to:

1. Build a dataset of gaze locations within news website interfaces.
2. Qualitatively evaluate SaRa (Saliency‐Driven Object Ranking), a saliency ranking framework proposed by Seychell and Debono which I am optimising, by examining its ability to assess the fairness of attention distribution across the content on these interfaces.

Check out updates on the project here.