Dynamic Facial Action Authentication for Anti-Spoofing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Face recognition-based identity authentication systems are vulnerable to low accuracy due to the ability of stolen photos or videos to pass authentication, as they do not require real-time participation from the authentication object.
Innovation Solution
The system provides prompt information to the authentication object to perform specific facial expressions or head actions, which are then matched against reference action information, ensuring that only authentic individuals with real-time participation can pass the authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If face recognition technology is used for identity authentication, then the authentication process is simple and fast, but the accuracy is low because stolen photos can pass authentication
Solution Approach 1:
The patent transforms static face recognition into dynamic action recognition by requiring the authentication object to perform specific facial expressions or head actions in response to prompts. The system captures and analyzes dynamic changes in facial features over time, making the authentication process resistant to static photos while maintaining speed through automated real-time analysis.
Solution Approach 2:
The system pre-establishes a library of reference action information containing expected facial expression changes and head action patterns for each user. During authentication, the system compares real-time captured actions against these pre-established references, enabling rapid verification while ensuring accuracy through predetermined action sequences that photos cannot replicate.
2Measurement precision
If action recognition is added to face recognition, then authentication accuracy is improved, but the device complexity increases
Solution Approach 1:
The patent merges face recognition and action recognition into a unified authentication system. The same imaging device captures both facial features and facial expressions/head actions, and the processing system integrates both recognition functions into a single authentication decision process, reducing overall system complexity while improving accuracy.
Solution Approach 2:
The imaging and processing system is designed to perform multiple functions: capturing static facial features for identification and capturing dynamic facial expressions/head actions for verification. This multi-functional approach eliminates the need for separate hardware systems, reducing complexity while maintaining high authentication accuracy.
Data Source
AI summary
Disclosed are an identity authentication method, an identity authentication device and a terminal. The identity authentication method includes: providing an authentication object with prompt information configured to prompt the authentication object to make a target authentication action, where the target authentication action includes at least one of a facial expression and a head action; collecting at least one frame of image; detecting whether authentication action information of the authentication object in the at least one frame of image matches reference action information generated by a target object making the target authentication action; and determining that the authentication object passes an identity authentication in a case that the authentication action information of the authentication object matches the reference action information.