A Systematic and Bibliometric Review on Face Recognition: Convolutional Neural Network

CNN (Convolution Neural Network) has boosted computer vision research by delivering outstanding outcomes in a variety of domains, including voice recognition, activity recognition, classification as well as segmentation of images, video analysis, as well as object recognition. In addition, the availability of cheap technology and massive volumes of data has created new possibilities for research for CNN. Alternative transfer functions, regularization, parameter optimization, including architectural enhancements, are a few inspiring ideas that have been researched for the advancement of CNN. Additionally, attaining architectural advancements leads to a significant increase in the deep CNN’s capability. This paper has a primary focus on a depth study of CNN (convolution neural network) on its activation functions and architecture for convolution neural networks. Furthermore, the bibliometric analysis has been done by using Scopus and Web of Science databases processed in VOSviewer software, by which researchers have ideas towards this domain. The findings of this paper will contribute to knowledge and practice of researchers working in this field by increasing their understanding of convolution neural networks.