A Review on Opportunities and Challenges of Machine Learning and Deep Learning for Eye Movements Classification

• Main Authors: Fikri, Muhammad Ainul, Santosa, Paulus Insap, Wibirama, Sunu
• Format: Conference or Workshop Item PeerReviewed
• Language: English
• Published: Institute of Electrical and Electronics Engineers Inc. 2021
• Subjects: Other Engineering; Engineering
• Online Access: https://repository.ugm.ac.id/281565/1/A_Review_on_Opportunities_and_Challenges_of_Machine_Learning_and_Deep_Learning_for_Eye_Movements_Classification.pdf; https://repository.ugm.ac.id/281565/; https://www.scopus.com/inward/record.uri?eid=2-s2.0-85124241626&doi=10.1109%2fIBITeC53045.2021.9649434&partnerID=40&md5=258e7f956280c2b40a88f4c3e5ddc6f7
• Institution: Universitas Gadjah Mada

Eye tracking has been used in touchless and assistive technologies to support disabled people as well as to provide more intuitive user interfaces. In this case, classification of events in eye tracking data is important to achieve higher object selection accuracy. Machine learning and deep learning have been used in events classification due to their ability to automatically learn patterns in eye tracking data. To the best knowledge of authors, however, there is no study that investigates opportunities and challenges on implementing various machine learning and deep learning techniques for events classification in eye tracking data. Here we present a systematical review to examine the use of machine learning and deep learning in events classification. We observed how machine learning and deep learning were used in development of reliable eye movements classification. At the same time, we summarized various challenges faced by previous researchers. In future, this paper may be used as a reference for entry level researchers interested in applying machine learning and deep learning for events classification in eye tracking data. © 2021 IEEE.