A rapid development in the banking industry for convolutional neural network-based signature verification
Ranadeep D. Reddy, Srinivasula A. Reddy, Mahesh Kotha, N. Suresh · 2025
A person’s signature is only a handwritten mark or sign that looks like their name, usually stylized and distinctive, and it signifies their identity, intent, and consent. Primarily used for many purposes, such as the authentication of legal documents, drafts, approvals, cheques, certifications, and correspondence. Since signatures are utilized in such crucial processes, it is crucial to verify their legitimacy. In the past, signatures were manually verified by comparing them to copies of authentic signatures. Given the rapid advancement of technology and the sophistication of signature forgeries and falsification procedures, this straightforward approach might not be adequate. The verification of handwritten signatures has been the subject of several research. Scholars have employed diverse methodologies to precisely discern legitimate signatures from expertly forged copies that bear striking resemblance to the authentic signature. Using many layers of hidden layers and receptive fields, signatures with unique properties can be correctly and quickly examined. This method’s primary contribution is to identify and reduce fraud, particularly in the banking sector.