Abstract
Texture analysis is extensively used for extraction of facial features. In this paper, we investigate extraction of facial features related to attributes of gender, age, race and expression. We propose novel approaches for texture analysis to improve single-label classification of these facial attributes. The proposed methods are derived by applying Local Binary Pattern based approaches on polar raster sampled face images. We perform experiments on three state-of-the-art face databases, namely, Face95, FERET and CK+. Experimental results show that the proposed approach improves the performance of Local Binary Pattern and its variants for single-label classification of facial attributes.