Liveness Detection via Eye Reflection Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video-based online services face challenges in authenticating users due to the inability to differentiate between real users and synthetic media, such as 'deep fake' simulations, which can impersonate genuine users, leading to potential security breaches.
Innovation Solution
A method and system that analyze reflections of a client device's visual content in reflecting surfaces, such as the user's eyes or glasses, to verify liveness by extracting and correlating reflection attributes, using dynamically changing visual objects and machine learning models to determine if the attributes are typical of human users or simulated.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If video-based authentication is used for online services, then user interaction becomes direct and easy, but the system becomes vulnerable to impersonation by malicious parties using synthetic media
Solution Approach 1:
The system displays visual objects with specific color patterns on the user's device screen and detects reflections of these colored objects in the user's eyes. The color information in the reflections serves as a liveness indicator, as synthetic media cannot accurately replicate the optical properties of human eyes. This resolves the contradiction by maintaining easy video-based interaction while adding color-based reflection analysis to prevent impersonation.
Solution Approach 2:
The system introduces an intermediary verification mechanism by analyzing reflections in the user's eyes as a mediator between the user and the authentication system. Instead of directly trusting video input, the system uses eye reflections as an intermediate layer of verification that confirms the presence of a real human user, thereby maintaining ease of use while enhancing security against synthetic media impersonation.
2Device complexity
If traditional face recognition is used for authentication, then the process is simple, but it cannot differentiate between real users and deep fake simulations
Solution Approach 1:
The system transitions from two-dimensional face recognition to three-dimensional reflection analysis by capturing and analyzing light reflections from the user's eyes. This adds a new dimension of verification that goes beyond flat facial images, enabling the system to detect liveness through optical properties that synthetic media cannot replicate, thereby improving detection precision while maintaining relatively simple implementation.
Solution Approach 2:
The system performs preliminary verification by analyzing eye reflections before completing the authentication process. By checking for the presence of real eye reflections containing visual object patterns early in the authentication flow, the system can quickly identify and reject deep fake attempts without requiring complex subsequent verification steps, thus maintaining simplicity while improving accuracy.
3Reliability
If complex verification methods are implemented to detect deep fakes, then authentication security improves, but computational resources required increase
Solution Approach 1:
The system extracts only the essential verification information from eye reflections - specifically, the presence and characteristics of reflected visual object patterns. By focusing on extracting this specific information rather than performing comprehensive analysis of the entire facial image, the system achieves reliable deep fake detection with minimal computational overhead, resolving the contradiction between security and resource consumption.
Solution Approach 2:
The system applies partial verification by analyzing only the eye reflection region rather than the entire facial image. This selective approach focuses computational resources on the most discriminative feature for liveness detection while avoiding unnecessary processing of other facial regions, thereby maintaining high authentication security with reduced computational resource requirements.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enhances the security and reliability of user authentication by significantly increasing the difficulty for synthetic media to simulate real-world reflection attributes, making it robust against sophisticated deep fake simulations, and reducing the computational resources required for verification.
Implementation Method 1
Receiving a sequence of images depicting one or more reflecting surfaces associated with the user viewing the screen while the one or more visual objects are displayed. Analyzing at least some images of the sequence to identify a reflection of at least part of the one or more displayed visual objects in at least part of the one or more reflecting surfaces.
Data Source
AI summary
Disclosed herein are methods and systems for determining liveness of a user, comprising analyzing visual content of a screen of a client device used by a user to access an online service, adjusting one or more visual objects displayed on the screen of the client device according to dynamically changing patterns, capturing a sequence of images depicting one or more reflecting surfaces associated with the user viewing the screen while the visual objects are displayed, analyzing the images to identify a reflection of the displayed visual objects in the reflecting surfaces and verifying liveness of the user based on one or more of a plurality of reflection attributes of the identified reflection.