Minimizing Error Rate in Gaze Tracking System based on Electrooculogram using Machine Learning

Abstract

Electrooculography is considered the most perceived signal processing system for distinguishing distinctive eye movements or eye-tracking. EOG signal is been utilized for extracting features for making reliable assistance for physically impaired patients. Extracting new features has been common and frequent in EOG studies. EOG system is less expensive than any other signal processing system and it incorporates disadvantages, which is the error rate. In our study, we calculated the Euclidean distance error and we got significantly less error rate, which is 4.02 cm. The principal objective of our study is to make a minimum error free EOG analysis. It is likewise less expensive than previous studies and this proposed strategy can be applied for continuous client experience for physically impaired patients inside a base expense.