FACIAL AGE-GROUP CLASSIFICATION WITH ORDINAL RANKING NEURAL NETWORK

Age-group classification is a technique for classifying a face image into a particular age group. However, due to uncontrollable environments, insufficient and incomplete training data, strong person-specific variations, and large within-age span variations, age-group classification has become a challenging problem. This paper presents a novel neural network with an ordinal ranking approach for automatic age-group classification. After pre-processing, histogram of oriented gradients (HOG) features are extracted. Then, the images are classified into age groups using an ordinal ranking neural network (ORNN) classifier. This classifier consists of a multiclass neural network binary classifier that categorizes the input images into different age groups. We experimented with this approach using four age groups derived from the FG-NET and MORPH-II datasets. ICIC International