Online Signature Recognition using Transformer Networks: An Overview

Online signature verification remains a pivotal component in digital security, ensuring the authenticity of digital transactions and documents. This paper offers a comprehensive overview of prevailing online signature verification techniques, with a particular emphasis on methods employing Deep Learning. RNNs, with their inherent capability to process sequential data, have been instrumental in improving the accuracy and efficiency of signature verification systems. However, as the digital forgery landscape becomes increasingly sophisticated, there is a pressing need to explore advanced neural architectures. In this context, the paper introduces a transformer-based approach for signature verification. Unlike traditional models, transformers, with their self-attention mechanisms, have the potential to capture intricate signature patterns and nuances, offering a promising avenue for future research and application