Facial Liveness Detection Using Device Motion Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing liveness detection methods for facial image data are inadequate in distinguishing between live and fraudulent biometric data, particularly in remote authentication transactions, leading to spoofing attacks that are difficult to detect.
Innovation Solution
An electronic device captures movement data during facial image capture, using a pre-trained machine learning algorithm to analyze this data and generate a confidence score, comparing it against a threshold to determine the authenticity of the facial image data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional liveness detection methods are used, then the authentication process is simple, but the detection accuracy is insufficient and spoofing attacks are difficult to detect
Solution Approach 1:
The patent applies dynamics by capturing movement data of the authentication device during the authentication process. Instead of using static images alone, the system analyzes dynamic movement patterns including device orientation changes, translation movements, and acceleration data to detect spoofing attempts, thereby improving detection accuracy while maintaining reasonable system complexity
Solution Approach 2:
The patent introduces movement data as an intermediary element between the facial image and the liveness detection decision. This intermediary provides additional information about the physical handling of the device during authentication, enabling more accurate detection of fraudulent attempts without requiring complex additional hardware
2Reliability
If movement data analysis is added to enhance liveness detection, then spoofing detection accuracy improves, but the processing complexity increases
Solution Approach 1:
The patent applies preliminary action by pre-processing and analyzing movement data during the authentication process itself. The system captures and evaluates device movement patterns in real-time before making the final authentication decision, allowing for reliable spoofing detection without requiring complex post-processing operations
Solution Approach 2:
The patent changes the parameters being analyzed by incorporating movement data parameters (device orientation, position, acceleration) alongside traditional facial recognition parameters. This multi-parameter approach enhances authentication reliability by providing multiple indicators to detect spoofing attempts
3Measurement precision
If multiple data types are captured and analyzed, then authentication accuracy improves, but the time required for authentication increases
Solution Approach 1:
The patent applies continuity of useful action by capturing movement data continuously during the authentication process rather than as separate discrete steps. The movement data collection occurs naturally throughout the authentication sequence, allowing for accurate analysis without adding significant time overhead to the overall process
Data Source
Figure 1
Figure 2~4
Figure 5
AI summary
A method for enhanced liveness detection of facial image data is provided that includes capturing movement data of an electronic device while capturing, using the electronic device, facial image data of a user. In response to determining the captured movement data is consistent with movement data expected to be captured during capture of facial image data, the method includes deciding the captured facial image data is genuine. In response to determining the captured movement data is different than movement data expected to be generated during capture of facial image data, the method includes deciding the captured facial image data is fraudulent.