Gaze-Based Authentication Using Front Camera for Continuous Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mobile device authentication methods, such as passwords and face recognition, are ineffective for ongoing user verification, especially when the device is already unlocked, and face recognition struggles with non-uniform illumination and appearance variations.
Innovation Solution
Implementing a gaze-based authentication system using a front-facing camera or other sensors to track and analyze eye movements, extracting features like gaze duration, scanning speed, and anticipatory gaze patterns to verify authorized users, even after initial device access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If password authentication is used, then device access can be controlled, but security is ineffective when device changes hands after unlocking
Solution Approach 1:
The system performs preliminary authentication by capturing the user's gaze pattern during the initial unlock process, storing it as a reference. This preliminary action enables subsequent continuous verification without requiring repeated passwords, solving the contradiction between authentication reliability and ease of operation.
Solution Approach 2:
The patent replaces the mechanical/password-based authentication system with an optical/biometric system using eye tracking. By substituting the manual password entry mechanism with automatic gaze pattern recognition, the system maintains high security reliability while eliminating the need for continuous manual verification.
2Reliability
If face recognition is used, then user verification can be performed, but it struggles with non-uniform illumination and appearance variations
Solution Approach 1:
The system extracts only the relevant iris and pupil regions from the entire face, isolating the gaze-specific features from other facial characteristics. This extraction approach makes the authentication system robust to illumination variations and appearance changes, as it focuses solely on the functional eye movement patterns rather than overall facial appearance.
Solution Approach 2:
Instead of analyzing the entire face uniformly, the system applies localized analysis to specific eye regions (iris, pupil, corneal reflection). This local quality approach allows the system to maintain high verification accuracy while being insensitive to global environmental conditions like lighting changes or facial expressions.
3Reliability
If gaze tracking is implemented for continuous authentication, then security is enhanced, but device complexity increases
Solution Approach 1:
The patent leverages the front-facing camera, already present for other device functions (selfies, video calls), to perform gaze tracking. By making the camera serve multiple purposes - both its original functions and biometric authentication - the system achieves continuous authentication without adding dedicated hardware, thus limiting the increase in device complexity.
Solution Approach 2:
The system uses the device's own existing camera and processor to perform gaze tracking and authentication, making the device authenticate itself through the user's natural gaze patterns. This self-service approach eliminates the need for external authentication hardware or complex additional systems, maintaining relatively simple device architecture while enabling continuous verification.
Data Source
AI summary
A method includes obtaining a gaze feature of a user of a device, wherein the device has already been unlocked using a second feature, the gaze feature being based on images of a pupil relative to a display screen of the device, comparing the obtained gaze feature to known gaze features of an authorized user of the device, and determining whether or not the user is authorized to use the device based on the comparison.