Design and Implementation on EMBA Authentication models

Authentication systems in general have faced long-term security deficit despite convoluted sophistication within various systems for decades. Unfortunately, humans serve as the weakest prey for exploitation within any security chain. Unskillful and inefficient communications between systems and humans give rise to multiple vulnerabilities related to any security model. Additionally, the Passwords/Personal Identification Numbers (PIN's) fail to provide idealistic security due to the lack of flawlessness in traditional password systems. On the contrary, since eye movements serve as a natural interaction modality, it was observed that the amalgamation of eye-tracking models enhances the overall security. In this context, the work presented in this paper leads to an extensive study and survey of many research works pertaining to existing Eye Movement authentication models. Besides, high-level overview of many conventional authentication methodologies have been discussed that can be launched in such EMBA models. Additionally, comparative analysis and performance metrics of different EMBA system is discussed. It was observed that the eye-password method has highest accuracy (97%) and eye motion based techniques has lowest accuracy (60%). Finally, an eye pupil tracking based authentication model has been proposed with accuracy of Eye detection, Eye open or closed detection and Eye pupil tracking detection are 98%, 92.51% and 96.25% respectively.