Human Age Estimation Using Facial Features

Abstract

Estimating human age automatically from the analysis of
the human face has lots of potential real-world applications, for
example human computer interaction (HCI) and multimedia
communication. But, even after developments in this field it is still
facial image classification. Age estimation is defined as a difficult
problem for the existing computer vision systems to effectively
estimate human ages and accurately. Age estimation is few of the
tasks of determination of a person's age or age group using facial
images. Results from several other experiments performed show that
face recognition and age estimation may be performed effectively
and can be achieved with low computational effort. A face detector
is learned by stage wise selection of the joint Haar-like features
using AdaBoost. In comparison to native implementation of
AdaBoost, fast AdaBoost reduces the time by approximately 50 to
100 times. With the introduction of Integral Image, features can be
computed quickly by the detector. Fast AdaBoost is an extension to
AdaBoost learning algorithm and it has proved to be a simple and
efficient classifier to select a less number of important visual
features from a huge set of promising features. The last and
foremost contribution is a method for combining classifiers in a
"cascade" which allows background areas in the image to be
discarded while achieving faster computation on promising face-like
regions.