Convolutional neural network-based face recognition using non-subsampled shearlet transform and histogram of local feature descriptors

<span lang="EN-US">Face recognition has been using in a variety of applications like preventing retail crime, unlocking phones, smart advertising, finding missing persons, and protecting law enforcement. However, the ability face techniques reduces substantially because changes pose, illumination, expressions individual. In this paper, novel approach based on non-subsampled shearlet transform (NSST), histogram-based local feature descriptors, convolutional neural network (CNN) is proposed. Initially, Viola-Jones algorithm used for detection then extracted region preprocessed by image resizing operation. Then, NSST decomposes input into low high-frequency component image. The descriptors such as phase quantization (LPQ), pyramid histogram oriented gradients (PHOG), proposed CNN are extracting features from low-frequency decomposition. fused to generate vector classified support machine (SVM). efficiency suggested method tested databases Olivetti Research Laboratory (ORL), Yale, </span><span lang="IN">J</span><span lang="EN-US">apanese female facial expression</span><span lang="IN">(</span><span lang="EN-US">JAFFE</span><span lang="IN">)</span><span lang="EN-US">. experimental outcomes reveal that outperforms some state-of-the-art approaches.</span>