Biometric Enrollment Sample Size Determination
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Biometric authentication systems face challenges in determining the adequate number of training samples required for accurate authentication performance, balancing security and user convenience, and efficiently evaluating the performance of collected samples.
Innovation Solution
The use of variance-based methods, including mathematical transformations and statistical analysis, to estimate the number of samples needed to achieve desired discriminative performance, allowing for dynamic adjustment of sample collection based on variability and performance metrics like FRR and FAR.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a large number of measurement samples are required for enrollment, then authentication accuracy is improved, but user convenience deteriorates
Solution Approach 1:
The patent implements dynamic sample size determination based on statistical evaluation of collected samples. The system continuously assesses the quality and representativeness of enrolled samples and adjusts the required sample size dynamically, allowing fewer samples when quality is high and more samples when quality is low, thus resolving the contradiction between accuracy and convenience
Solution Approach 2:
The patent changes the parameter of sample size from a fixed requirement to a variable value determined by statistical evaluation. By using metrics like FRR and FAR to evaluate sample quality and adjusting the number of required samples accordingly, the system achieves both high authentication accuracy and improved user convenience
2Measurement precision
If the permissible variability in measurements is reduced, then authentication accuracy is improved, but the false reject ratio increases
Solution Approach 1:
The patent introduces feedback mechanisms through statistical evaluation of collected samples. The system continuously monitors authentication performance metrics (FRR and FAR) and uses this feedback to adjust the enrollment process, ensuring that the template achieves the desired balance between accuracy and reliability by adapting to actual measurement characteristics
Solution Approach 2:
The patent performs preliminary statistical evaluation and template construction before actual authentication occurs. By pre-processing and evaluating samples to determine optimal template characteristics, the system establishes accurate thresholds that maintain both low false accept ratios and acceptable false reject ratios, preventing accuracy-reliability trade-offs during operation
3Measurement precision
If formal methods of evaluating samples are implemented, then determination of adequate sample size is improved, but system complexity increases
Solution Approach 1:
The patent implements self-service evaluation where the system automatically assesses its own sample quality using built-in statistical methods. The evaluation process is integrated into the enrollment workflow, allowing the system to self-determine when sufficient samples have been collected without requiring external intervention or complex manual evaluation procedures, thus managing complexity while improving precision
Data Source
AI summary
At least two biometric measurements of a person are collected, then a statistical measure based on the measurements is computed. The statistical measure is a bounded estimate of the discriminative power of a test based on the measurements. While the discriminative power is less than a target value, additional biometric measurements are collected. When enough measurements have been collected, a biometric template is constructed from the measurements and stored for use in future identifications. Systems and software to implement similar methods are also described and claimed.