Occluded face recognition based on the improved SVM and block weighted LBP

Facial occlusions, for example, sunglasses, and scarves, etc., can significantly affect the performance of any facial recognition system. The focus of this paper is on facial occlusions, and particularly, on how to improve the recognition of faces occluded by sunglasses and scarves. We propose a new approach that consists of first detecting the presence of sunglasses/scarves and then processing the non-occluded facial regions only. The occlusion detection problem is approached using PCA and improved support vector machines (SVM), while the recognition of the non-occluded facial part is performed using blocked-based weighted local binary patterns (LBP).