Dynamic Feature Subset Selection for Occluded Face Recognition

Accurate recognition of person identity is a critical task in civil society for various application and different needs. There are well-established biometric modalities that can be used purposes such as face, voice, fingerprint, iris, etc. Recently, face images have been widely recognition, since the human most natural user-friendly method. However, real-life applications, some factors may degrade performance, partial occlusion, poses, illumination conditions, facial expressions, In this paper, we propose two dynamic feature subset selection (DFSS) methods to achieve better occluded faces. The proposed DFSS based on resilient algorithms attempting minimize negative influence confusing low-quality features extracted from areas by either exclusion or weight reduction. Principal Component Analysis Gabor filtering approaches initially extraction, then dynamically enforced. This leading more effective representation an improved performance. To validate their effectiveness, multiple experiments conducted performance compared. experimental work carried out using AR database Extended Yale Face Database B. obtained results identification verification show both outperform standard (static) use whole number equal weights.