Age estimation based on head movements: A feasibility study

Abstract

The ability to estimate the age of a machine user is a highly important task that can be used as the basis for developing age adaptive human computer interaction systems. Recently a considerable number of researchers described work on facial age estimation, where facial features are used for estimating the age of a subject. Usually in facial age estimation systems, static images are used so that the dynamics of face movements are not taken into consideration during the process of age estimation. In this paper we present a framework for age estimation based on head movements. The feasibility of the proposed framework is evaluated using a dedicated experimental procedure where we compare the performance of age estimation using head movements against age estimation based on mouse movements and age estimation based on face images. Preliminary results demonstrate the potential of using head motion as the basis for estimating the age of machine users.