A new classification method based on pairwise SVM for facial age estimation

Abstract

This paper presents a practical algorithm for facial age estimation from frontal face image. Facial age estimation generally comprises two key steps including age image representation and age estimation. The anthropometric model used in this study includes computation of eighteen craniofacial ratios and a new accurate skin wrinkles analysis in the first step and a pairwise binary support vector machine (SVM) in the second one. Anthropometric model is the first model that has been provided; however, it hasn't been much considered and even hasn't been applied on any large database so far. Therefore, the algorithm is applied on FG-Net database and the average of the absolute errors (MAE) and cumulative score (CS) measures are provided to make comparison with other approaches much easier. Experimental results show that the proposed method can give MAE=6.34 and CS (<=10) =81.14 using a pairwise binary tree support vector machine (SVM).