Gorde:
Xehetasun bibliografikoak
Egile nagusia: Patil, Aditya Lotan
Formatua: Recurso digital
Hizkuntza:ingelesa
Argitaratua: Zenodo 2025
Gaiak:
Sarrera elektronikoa:https://doi.org/10.5281/zenodo.17480968
Etiketak: Etiketa erantsi
Etiketarik gabe, Izan zaitez lehena erregistro honi etiketa jartzen!
Aurkibidea:
  • <p><span>This paper proposes a </span><strong><span>Secure Authentication Framework</span></strong><span> that integrates </span><strong><span>Java Spring Boot</span></strong><span> and </span><strong><span>AI-driven Facial Recognition</span></strong><span> to achieve robust multi-factor authentication (MFA)</span></p> <p> </p> <div></div> <p></p> <div></div> <p></p> <p></p> <p> </p> <p><span>The architecture uses a </span><strong><span>polyglot microservices</span></strong><span> design, leveraging Spring Boot for the secure backend and business logic, and a </span><strong><span>Python service (DeepFace/VGG-Face)</span></strong><span> for high-accuracy facial analysis<sup></sup><sup></sup><sup></sup><sup></sup><sup></sup><sup></sup><sup></sup><sup></sup><sup></sup></span>.</p> <div></div> <p></p> <div></div> <p></p> <div></div> <p></p> <p> </p> <p>Key contributions include:</p> <ul> <li> <p><span>A multi-step login flow combining password checking with a biometric factor<sup></sup></span>.</p> <div></div> <p> </p> </li> <li> <p><span>The implementation of </span><strong><span>Liveness Detection</span></strong><span> to defeat spoofing attacks (photos/videos)<sup></sup><sup></sup><sup></sup><sup></sup></span>.</p> <div></div> <div></div> <p> </p> </li> <li> <p><span>Enhanced privacy by storing only a non-recreatable </span><strong><span>facial embedding</span></strong><span> (string of numbers) rather than the user's raw photo<sup></sup><sup></sup><sup></sup><sup></sup></span>.</p> <div></div> <div></div> <p> </p> </li> <li> <p><span>A detailed blueprint for a scalable system where the AI service is </span><strong><span>stateless</span></strong><span> and isolated from the user database, significantly reducing the attack surface<sup></sup><sup></sup><sup></sup><sup></sup></span>.</p> <div></div> <div></div> <p> </p> </li> </ul> <p><span>The conceptual results demonstrate a high level of security, achieving a low </span><strong><span>False Acceptance Rate (FAR)</span></strong><span>, which is critical for protection against cyber threats<sup></sup></span></p>