Texture-based Estimation of Age and Gender from Wild Conditions

Abstract

The paper concerns the estimation of facial attributes—namely, age and gender—from images of faces acquired in challenging, in the wild conditions. This problem has received far less attention than the related problem of face recognition, and in particular, has not enjoyed the same dramatic improvement in capabilities demonstrated by contemporary face recognition systems. Here, this problem is addressed by making the following contributions. First, in answer to one of the key problems of age estimation research—absence of data—a unique data set of face images, labelled for age and gender is offered, acquired by smart-phones and other mobile devices, and uploaded without manual filtering to online image repositories. The images in this collection are more challenging than those offered by other face-photo benchmarks. Second, a dropout-support vector machine approach is described used by this system for face attribute estimation, in order to avoid overfitting. Inorder to make classification of age using kNN more easy, texture features are extracted. Finally, a robust face alignment technique is presented, which explicitly considers the uncertainties of facial feature detectors.