Face Authentication System Using Movement Continuity Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current information processing systems that authenticate individuals using face images are vulnerable to spoofing, where a malicious individual can pretend to be a registered person by displaying their image on a screen, leading to false authentication success.
Innovation Solution
The system continuously images a target and extracts feature quantities that change with the target's movement, determining authentication failure when these features change discontinuously, thereby preventing spoofing by ensuring continuity in movement patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If authentication is performed using a single face image, then authentication speed is fast, but the system is vulnerable to spoofing attacks
Solution Approach 1:
The system performs preliminary imaging of the target's face in multiple images before authentication decision is made. By capturing multiple images in advance and storing them, the system can later analyze movement continuity without requiring real-time complex processing during the actual authentication moment, thus improving security while managing complexity.
Solution Approach 2:
The patent applies continuity by analyzing whether the target's face movement is continuous across multiple images. The authentication unit checks if the extracted feature quantities show continuous change patterns, which is a characteristic of real human faces. This continuity check adds a temporal dimension to authentication, making spoofing difficult while maintaining a relatively simple implementation.
2Reliability
If multiple images are captured and movement continuity is analyzed, then spoofing resistance is improved, but authentication time increases
Solution Approach 1:
Multiple images are captured and stored in advance before the authentication decision is needed. This preliminary image collection allows the system to perform movement continuity analysis without delaying the actual authentication moment, as the images are already ready for processing.
Solution Approach 2:
The system extracts only the necessary feature quantities from the multiple captured images that are relevant for movement continuity analysis. By selecting and extracting only essential features rather than processing all image data, the system reduces computational time while maintaining effective anti-spoofing capability.
3Measurement precision
If feature quantities are extracted from multiple images, then authentication accuracy is improved, but processing complexity increases
Solution Approach 1:
The system extracts only the necessary feature quantities from the multiple captured images that are relevant for movement continuity analysis. By selecting and extracting only essential features rather than processing all image data, the system reduces computational complexity while maintaining high authentication accuracy.
Solution Approach 2:
The patent focuses the feature extraction on specific local characteristics of the face that change with movement, such as facial feature positions or expressions. By concentrating on locally relevant features rather than analyzing the entire image, the system improves accuracy for movement detection while reducing overall processing complexity.
Data Source
AI summary
An information processing system includes: an imaging apparatus that generates a plurality of images, by imaging the target continuously in time series; and an information processing apparatus that authenticates the target by using at least one of the plurality of images. The information processing apparatus includes: an extraction unit that extracts, from each of the plurality of images, a feature quantity of target that changes in accordance with a movement of the target; and an authentication unit that determines that authentication of the target is failed, when the feature quantity changes discontinuously in time series among the plurality of images.