Human Activity Recognition using Eye Movements: A review

Human activity recognition represents a prominent and formidable challenge within the domain of computer vision. This problem holds substantial relevance across various domains, encompassing robotics, gaming, therapeutic interventions, human-robot interaction, athletic performance analysis, healthcare monitoring, and video surveillance. Due to the complicated human pose variations and group interactions, human activity recognition is difficult. Some of these inherent difficulties could be solved using sensors with several modalities. An RGB-D camera, thermal sensors, infrared sensors etc. might all be examples of such sensors. Due to intrusive nature of these sensing modalities it may cause user inconvenience which limits its utility to a broader audience. To overcome this, in this work, we explore eye movement analysis for sensing activity recognition. Eye movement analysis has become a promising method for identifying human activities across a variety of fields. This paper provides a review of techniques and applications for eye movement analysis in human activity recognition. The study examines the possibility of detecting human activity using a low-cost, camera based eye gaze tracking device and opens up discussion on the wider applicability of eye movements based activity recognition that are difficult, or even impossible, to detect using common sensing modalities.