Individualized Biometric Threshold Determination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current biometric authentication systems face challenges in determining individual thresholds for false matching rates, leading to inconsistent verification accuracy and requiring multiple enrollment processing verifications, which are time-consuming and inconvenient.
Innovation Solution
A threshold determining device comprising an individual threshold evaluation unit, an individual threshold distribution evaluation unit, and a calculation unit to determine specific thresholds for each data set, ensuring a prescribed false matching rate is met, allowing for individualized threshold determination and overall threshold calculation based on distribution analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If a fixed threshold is used for verification, then the authentication process is simple, but the false matching rate and false non-matching rate cannot be controlled for each individual data set
Solution Approach 1:
The patent changes the threshold parameter from a fixed value to individualized values calculated for each data set. The threshold calculation unit computes specific threshold values based on verification evaluation values and false matching rate requirements for each individual data set, thereby achieving both reliability (controlled false matching rates) and adaptability while maintaining operational simplicity through automated calculation.
Solution Approach 2:
The patent segments the single fixed threshold into multiple individual thresholds, one for each data set. By dividing the universal threshold into data-specific thresholds, the system can control false matching rates individually for each data set while maintaining overall system reliability. This segmentation allows each data set to have its own optimized threshold rather than using a one-size-fits-all approach.
2Reliability
If individual thresholds are determined through verification experiments at enrollment, then the false matching rate for each individual can be controlled, but the enrollment processing becomes time-consuming and complex
Solution Approach 1:
The patent performs preliminary action by pre-calculating individual thresholds using the threshold calculation unit before actual authentication occurs. Instead of conducting time-consuming verification experiments during enrollment, the system pre-computes thresholds based on verification evaluation values and required false matching rates, storing them for later use. This eliminates the need for extensive real-time verification during enrollment while maintaining reliability.
Solution Approach 2:
The patent replaces the mechanical verification experiment process with an automated threshold calculation mechanism. Instead of physically conducting multiple verification tests during enrollment to determine thresholds, the system uses the threshold calculation unit to compute thresholds mathematically based on verification evaluation values and false matching rate requirements, significantly reducing enrollment time and complexity.
3Device complexity
If a fixed threshold is used, then the authentication system is easy to implement, but the average false matching rate cannot be reduced below a certain level
Solution Approach 1:
The patent changes the threshold parameter from a fixed system-wide value to individually calculated values for each data set. The threshold calculation unit computes specific threshold values based on verification evaluation values and false matching rate requirements, enabling the system to achieve lower average false matching rates while maintaining manageable complexity through automated calculation processes.
Solution Approach 2:
The patent implements feedback by using the false matching rate requirement as input to the threshold calculation unit, which then determines appropriate individual thresholds. This feedback mechanism ensures that the calculated thresholds will achieve the desired false matching rate performance. The system continuously optimizes thresholds based on verification evaluation values and required performance levels, maintaining reliability while controlling complexity.
Data Source
AI summary
Provided is a threshold determining device for determining a threshold, at which the false matching rate of each data to become less than a required value becomes a required assurance value or more. The threshold determining device comprises an individual threshold evaluation unit, an individual threshold distribution evaluation unit, and an overall threshold determination unit. The threshold determining device determines a similarity of each data to others for each data, determines the threshold of the similarity satisfying a predetermined false matching rate, for each data, determines an individual threshold distribution of the data for each threshold on each of the data, and determines the threshold common to the entire data, as overall data on the basis of the individual threshold distribution.