Authors :
Dr. Praveen Kumar K V; Pramit Kumar Mandal; Rishav Anand; Sakshee Singh; Tsewang Choskit
Volume/Issue :
Volume 9 - 2024, Issue 3 - March
Google Scholar :
https://tinyurl.com/2my92p96
Scribd :
https://tinyurl.com/5c8bsh9z
DOI :
https://doi.org/10.5281/zenodo.10793129
Abstract :
The usage of advanced signature verification
technologies is required because of the growing
dependence on digital transactions and authentication
technology. This survey looks at the current state of
dynamic signature representation techniques, with a
focus on learning without forgeries. The efficacy of
enhancing the security of signature-based authentication
systems through the combination of 1D CNNs and the
novel signature embedding approach Synsig2Vec is
assessed.
The survey's first section addresses the dangers of
forgery attacks and the weaknesses of employing
traditional signature verification methods. It then
explores the state- of-the-art Synsig2Vec methodology,
which provides a more thorough representation by
capturing the dynamic characteristics of signatures. By
adding 1D CNN, the feature extraction procedure is
further enhanced and the model's accuracy in
differentiating real signatures from fakes is increased.
Keywords :
Signature Verification, Forgery Detection, Synsig2Vec, 1D CNN, Dynamic Signature Representation, Authentication Systems.
The usage of advanced signature verification
technologies is required because of the growing
dependence on digital transactions and authentication
technology. This survey looks at the current state of
dynamic signature representation techniques, with a
focus on learning without forgeries. The efficacy of
enhancing the security of signature-based authentication
systems through the combination of 1D CNNs and the
novel signature embedding approach Synsig2Vec is
assessed.
The survey's first section addresses the dangers of
forgery attacks and the weaknesses of employing
traditional signature verification methods. It then
explores the state- of-the-art Synsig2Vec methodology,
which provides a more thorough representation by
capturing the dynamic characteristics of signatures. By
adding 1D CNN, the feature extraction procedure is
further enhanced and the model's accuracy in
differentiating real signatures from fakes is increased.
Keywords :
Signature Verification, Forgery Detection, Synsig2Vec, 1D CNN, Dynamic Signature Representation, Authentication Systems.
