Eye Movement and Visual Target Synchronization Level Detection Using Deep Learning

In recent years, deep learning has been widely used in the eye-tracking area. Eye-tracking has been studied to diagnose neurological and psychological diseases early since it is a simple, non-invasive, and objective proxy measurement of cognitive function. This project aims to develop a system to automatically track the synchronisation of eye movement data and its visual target. To achieve this goal, we employ a deep learning algorithm (Points-CNN and Head-CNN) to detect the eye centre location and classify the synchronisation level of the eye movement and visual target. Moreover, we modify the eyediap dataset to assist with our research objective. The video data in the eyediap dataset is used to track the eye movement trajectory, while the visual target movement data is used to extract the direction change window. The movement feature vectors are extracted from the eye movement data and the visual target movement data with the direction change window. Euclidean distance, Cosine similarity, and Jaccard similarity coefficient are used to assist the synchronization detection of the eye and visual target movement vector. In the synchronisation detection part, K-Nearest Neighbors, Support Vector Machine, Logistic Regression are investigated.