Comparison of Signature Forgery Detection Architectures
E. Surya Teja · International Journal for Research in Applied Science and Engineering Technology · 2022
Abstract: In today's society, we use signatures for many important documents like passports, driving licenses, bank cheques, etc. But signatures can be forged in multiple ways, which can create a number of problems, such as identity theft, hacking, fake identification, etc. To reduce these problems, our project is about developing a system for detecting whether a signature is forged or real from a dataset of signatures. For our project, we are developing an offline signature verification system. This system is based on CNN (Convolutional Neural Network) and SNN (Siamese Neural Network). In this project, we will be comparing both CNN’s and SNN's to find which produces a better result. We are implementing the project using a custom CNN, a CNN with VGG16 architecture, a custom SNN and a SNN using the SigNet architecture.