Adaptive Dual-Factor Biometric Identification System
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Biometric identification systems face authorization errors due to statistical variations in biometric identifying information during enrollment and authentication, leading to unreliable unique authentication identifiers and increased authentication denials.
Innovation Solution
An adaptive dual-factor identification system that updates enrolled users' unique authentication identifiers based on biometric measurements provided during authorization requests, using two Deep Neural Networks (DNNs) to compute embeddings for facial and vocal features, and adjusts these identifiers based on confidence levels to improve authentication accuracy over time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If biometric identifying data is collected during enrollment to compute authentication identifiers, then authentication capability is established, but statistical variations in the biometric data cause unreliable identifiers and authorization errors
Solution Approach 1:
The system dynamically updates enrollment embeddings during authentication attempts. When a authentication attempt occurs, the system computes an authentication embedding and uses it to update the enrollment embedding region, making the enrollment data adaptive and dynamic rather than static. This allows the system to adapt to natural variations in biometric data over time.
Solution Approach 2:
The system implements a feedback mechanism where authentication embeddings are fed back into the enrollment database to refine future authentication decisions. The authentication embedding computed during a authentication attempt is used to update the enrollment embedding region, creating a closed-loop system that continuously improves its accuracy based on actual authentication outcomes.
2Reliability
If enrollment embedding regions are computed using totality of biometric points, then comprehensive user representation is achieved, but outlier embeddings and statistical variations cause enrollment regions to exclude valid user embeddings
Solution Approach 1:
Instead of requiring all biometric points to define the enrollment region, the system uses a partial approach by computing the enrollment embedding region based on authentication embeddings. This allows the system to focus on the most relevant authentication outcomes rather than being constrained by all collected biometric data, including outliers.
Solution Approach 2:
The system changes the parameters used to define enrollment regions from static biometric point collections to dynamic authentication embeddings. By computing enrollment regions based on authentication outcomes rather than raw biometric data, the system transforms the parameter basis from potentially noisy raw data to refined authentication results.
Data Source
AI summary
A multi-factor identification system is provided in which enrolled user authentication information is updated in the course of an authorization request based upon at least one of a confidence level of a match between a request first factor identifier, produced based upon first unique user identifying information received with the authentication request, and a respective matching enrolled first factor identifier and a confidence level of a match between a request second factor identifier, produced based upon second unique user identifying information received with the authentication request, and a respective matching enrolled second factor identifier.