Model-based head pose-free gaze estimation for assistive communication

Abstract

The significance of employing video-based eye-gaze tracking as an assistive tool has long been recognised, especially in the domain of human–computer interaction to assist physically challenged individuals in operating a computer by the eye movements alone. Nonetheless, several operating conditions typically associated with existing eye-gaze tracking methods, relating to constraints on the head movement and prolonged user-calibration prior to gaze estimation, need to be alleviated in order to better assist individuals with motor disabilities. In this paper, we propose a method that is based on a cylindrical head and spherical eyeballs model to estimate the three-dimensional eye-gaze under free head movement from a single camera integrated into a notebook computer, alleviating any assumptions of stationary head movement without requiring prolonged user co-operation prior to gaze estimation. The validity of the proposed method has been investigated on a publicly available data set and real-life data captured through the voluntary collaboration of a group of normal subjects and a person suffering from cerebral palsy.