Facial Recognition Anti-Spoofing via Proximity Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Facial recognition technologies are vulnerable to spoofing attempts, where unauthorized users can gain access by presenting images or videos of authorized users, leading to erroneous authentication and resource wastage.
Innovation Solution
Implementing anti-spoofing techniques that use proximity sensors and image analysis to determine the actual distance between the device and the object in the image, denying authentication if the object is too close or if the image lacks sufficient pixels between facial features, thereby preventing erroneous access.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If facial recognition is performed on captured images, then user authentication is enabled, but computing resources and battery life are wasted when spoofing attempts are detected
Solution Approach 1:
The system performs preliminary distance measurement using the image capture device before conducting facial recognition. By detecting whether the object is within a threshold distance, the system prevents unnecessary facial recognition processing in cases of obvious spoofing attempts, thereby conserving battery life while maintaining security.
Solution Approach 2:
The patent introduces an intermediary distance detection mechanism that acts as a filter between the authentication request and the facial recognition processing. This intermediary step evaluates the spatial relationship between the device and the presented object, mediating whether full authentication processing should proceed.
2Reliability
If distance detection is performed before facial recognition, then spoofing attempts are prevented, but device complexity increases
Solution Approach 1:
The image capture device serves multiple functions: it captures images for facial recognition and simultaneously measures distance for spoofing detection. By making the capture device universal, the patent avoids adding separate hardware components, thereby limiting the increase in device complexity while achieving anti-spoofing capability.
Solution Approach 2:
The patent merges the distance measurement function with the existing image capture functionality. Instead of implementing separate distance detection hardware, the system combines both image capture and distance measurement into a single integrated process using the same capture device.
3Measurement precision
If facial recognition processing is performed on all images, then authentication accuracy is maintained, but computing resources are wasted on obvious spoofing attempts
Solution Approach 1:
The system performs a preliminary distance check before initiating facial recognition processing. This preliminary action filters out obviously spoofed images (where the object is too close) before they enter the processing pipeline, maintaining authentication accuracy for legitimate users while improving overall processing efficiency by avoiding wasted computation on obvious fakes.
Data Source
AI summary
An example method includes capturing, by an image capture device of a computing device, an image of a face of a user. The method further includes detecting, by the computing device, whether a distance between the computing device and an object represented by at least a portion of the image is less than a threshold distance, and, when the detected distance is less than a threshold distance, denying authentication to the user with respect to accessing one or more functionalities controlled by the computing device, where the authentication is denied independent of performing facial recognition based at least in part on the captured image.