Multimodal Biometric Feature Binding for Secure Enrollment and Authentication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Automated multi-modal biometrics systems lack security measures to ensure that biometric features acquired from different biometric sensors belong to the same individual, particularly in the absence of human supervision, making them vulnerable to spoofing attacks.
Innovation Solution
A method for enrollment or authentication in multi-modal biometrics systems that involves acquiring first and second biometric data, deriving corresponding biological features, and determining a similarity score to verify if these features originate from the same person, using a classification model to enhance security.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If automated multi-modal biometrics systems are deployed without human supervision, then convenience and efficiency are improved, but security is worsened due to inability to verify that biometric features belong to the same individual
Solution Approach 1:
The patent introduces an intermediary verification mechanism that compares temporal patterns and spatial relationships of multiple biometric features to determine if they belong to the same individual. This intermediary check acts as a mediator between automated acquisition and security verification, enabling automated systems to enforce identity binding without human supervision.
Solution Approach 2:
The system implements feedback by analyzing the temporal sequence and spatial configuration of acquired biometric features, then using this feedback to verify whether multiple features originate from the same individual. The verification result feeds back into the enrollment/authentication decision, allowing automated security enforcement.
2Speed
If multiple biometric features are acquired in parallel or sequentially without verification, then enrollment speed is improved, but measurement precision is worsened regarding identity attribution
Solution Approach 1:
The patent applies preliminary action by capturing temporal and spatial metadata during the biometric acquisition process itself, before verification is needed. The system records timing information and spatial relationships as biometric features are acquired, preparing verification data in advance without slowing down the acquisition process.
Solution Approach 2:
The system adds another dimension to biometric verification by incorporating temporal sequencing and spatial positioning information alongside traditional biometric features. This dimensional expansion allows the system to verify identity attribution without requiring slower, sequential verification steps.
3Adaptability or versatility
If biometric features are acquired from different sensors without verification, then device versatility is improved, but reliability is worsened due to spoofing attacks
Solution Approach 1:
The patent implements universality by creating a verification mechanism that works across multiple biometric modalities and sensor types. The temporal-spatial analysis approach is modality-agnostic, enabling the same verification principle to protect fingerprint, facial, iris, and other biometric sensors uniformly against spoofing attacks.
Solution Approach 2:
The system applies preliminary anti-action by proactively verifying the authenticity of biometric features before they are used for enrollment or authentication. The temporal-spatial consistency check prevents spoofed features from being accepted, countering potential attacks before they can compromise the system.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention provides a method for enrollment or authentication of an individual in a multi-modal biometrics system comprising at least one biometrical sensor, the multi-modal biometrics system being configured to acquire first biometrical data representative of a first biometrical feature and second biometrical data representative of a second biometrical feature, via the at least one biometrical sensor. The method comprises the following steps: - obtaining (203) a first biological feature based on first data acquired by the biometrical sensor acquiring the first biometrical data; - obtaining (204) a second biological feature based on second data acquired by the biometrical sensor acquiring the second biometrical data; - determining (205) a similarity score of the first biological feature and the second biological feature; - based on the similarity score, determining (206) whether the first biological feature and the second biological feature are from the same person or not.