Abstract

There are 3 state of the art approaches for recognizing faces under varying poses. These 3 approaches are Overlapping Discrete Cosine Transform (DCT), Hybrid Spatial Feature Interdependence Matrix (HSFIM) and Score Level Fusion Techniques (SLFT). The train and test images are considered from standard public face data bases: Head Pose Image Database, Sheffield Face Database, and Indian Face Database. The key contribution of this article is, we have developed and analyzed the 3 state of the art approaches for recognizing faces under varying poses using a common set of train and test images. This evaluation gives us the exact face recognition rates of the 3 systems under varying poses. We have considered patents ‘Face recognition apparatus, face recognition method, gabor filter application apparatus, and computer program’ ‘Gabor filtering and joint sparsity model-based face recognition method’ ‘Face identification method based on multiscale weber local descriptor and kernel group sparse representation’.

Keywords: Discrete cosine transform, face recognition, hybrid spatial feature interdependence matrix, score level fusion.