Intelligent Secure Authentication of Smart Devices Using Eye Tracking: Design and Implementation

Pattern-based, password-based, and PIN-based authentication techniques are typically the most common among smartphone users. However, according to prior studies these authentication techniques are susceptible to keylogger, smear, and shoulder surfing assaults. These attacks take advantage of data from the user’s handwriting or finger movements. Some research studies restrict eye-tracking authentication for smartphone users because of the regularly controlled environmental circumstances. Currently, no smartphone models support eye-tracking-based authentication. This paper introduces a novel eye-tracking user identification technique using a machine-learning model with inception modules and an ensemble structure for smartphones that eliminates the need for unique hardware or calibration procedures. The goal is to have a favorable influence on the development of an eye-tracking-based method for smart device user verification. The eye movement trajectories as the input to build a character for authentication. These are recorded by the front camera of a smartphone and show the user’s eye movement direction. The proposed application would assist users in situations where they can’t touch their screens for various reasons, such as dirty or stained hands. The proposed method reduces the preprocessing steps typically used in conventional methods. A prototype of the security-based application would be constructed to test the model on a smartphone.