A hybrid SUGWO optimization for partial face recognition with new similarity index

This paper introduces a Partial Face Recognition (PFR) method with the benefits of optimization logic using an optimized feature matching aspect. Besides, for better recognition, the Sparse Representation Classification (SRC) and Fully Convolutional Network (FCN) have been combined. As a novelty, this work aims to tune the sparse coefficient of Dynamic Feature Matching (DFM) optimally, in which the reconstruction error should be minimal. Also, this work presents the structural similarity index measure to calculate the similarity scores between the gallery sub-feature map and probe feature map. For optimization purposes, this work deploys a proposed Sealion Updated Grey Wolf Optimization (SUGWO) algorithm. Finally, the proposed method is executed over the traditional methods concerning certain measures.