55 Hierarchical Face Age-Estimation Algorithm Using Informed Facial Features

This paper introduces a novel technique for estimating the age of a person from a digital image of the face. The age estimating technique proposed combines Active Appearance Models (AAMs) and machine learning methods, i.e. Artificial Neural Network (ANN) and Support Vector Regression (SVR), to improve the accuracy of human age estimation over the current state-of-the-art algorithms. In this method, characteristics of the face are codified into feature vectors by the use of a multi-factored Principle Components Analysis (PCA) as utilized by AAMs. The feature vectors are provided as input to the ANN for a binary group classification: youth and adult. A unique age estimation function is derived for each group using SVR of the feature vector. The proposed approach yields significant improvement in overall mean-absolute error (MAE), mean-absolute error per decade of life (MAE/D), and the Percent Error Cumulative Score (CS) against the baseline data corpus.