A Robust Face Recognition Method for Expression and Pose Variant Images

Face Recognition (FR) has been vastly studied in the past few decades and researchers have been able to achieve success in this field, mostly in controlled environments. However, the FR algorithm faces huge drop in its accuracy due to the variation in pose and expression in face images. This paper proposes a novel FR method to overcome the issues of pose and expression. The proposed FR algorithm initially locates faces using dual shot face detection algorithm. Later, an intelligent combination of Linear Discriminant Analysis (LDA) and Adaptive Boosting is used to extract features from detected and normalized face images. Finally, classic nearest center classifier is applied to achieve classification. Detailed experiments are conducted on JAFFE and KDEF databases. Moreover, a thorough analysis and comparison of the effect of pose and expression variation on the accuracy of FR is presented. The experimental results show that the proposed algorithm achieves better recognition accuracy than the recent FR methods.