Development of signature recognition system using VGG16

Abstract

The goal of this research is to locate a signature recognition system that can replace individual presence in the process of handwritten signature recognition. It has been discovered that an automatic signature recognition system that does not require a person’s physical presence is necessary throughout corona duration. In this article more accurate signature recognition model employing VGG 16 pre- trained models is proposed. Novel Convolution neural network has exhibited 83% validation accuracy on the GPDS synthetic Signature dataset[1].