Multi-modal Biometric Fusion via Boosted Classifier Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current user authentication methods, particularly in mobile and wearable devices, face challenges in providing robust and efficient multi-modal authentication solutions that effectively combine different biometric modalities such as face, fingerprint, and iris recognition, often resulting in inconveniences and security risks due to the limitations of individual recognition technologies.
Innovation Solution
A multi-modal fusion method that processes input images from various biometric modalities using classifiers based on boosting schemes to determine fusion parameters, allowing for weighted summation of scores from different modalities for enhanced authentication, thereby improving security and convenience.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple biometric modalities are combined for authentication, then security and reliability are improved, but system complexity increases
Solution Approach 1:
The patent segments the authentication system into independent modality-specific classifiers (e.g., face classifier, fingerprint classifier, iris classifier) that process each biometric modality separately. Each classifier operates independently on its own input image, producing individual scores that are then combined through a fusion mechanism. This segmentation allows the system to maintain high reliability through multi-modal integration while managing complexity by keeping each classifier focused and specialized.
Solution Approach 2:
The patent merges the outputs of multiple independent classifiers through a score fusion mechanism. The fusion process combines scores from different biometric modalities (face, fingerprint, iris, etc.) to produce a final authentication decision. This merging approach enables the system to leverage the strengths of multiple modalities, improving overall reliability and security while maintaining a structured framework that manages system complexity.
2Measurement precision
If multiple classifiers are used for different biometric modalities, then authentication accuracy is improved, but processing time increases
Solution Approach 1:
The patent divides the authentication process into parallel segments, where each modality-specific classifier processes its input image independently and simultaneously. Rather than processing modalities sequentially, the system executes multiple classifiers in parallel, which reduces the cumulative processing time while maintaining high authentication accuracy through the combined results of all classifiers.
Solution Approach 2:
The patent performs preliminary processing of input images for each modality before classification, including tasks such as image normalization, feature extraction, and preprocessing specific to each biometric type. This preliminary action prepares the data in advance, allowing the actual classification process to be faster and more efficient, thereby reducing overall processing time while maintaining high accuracy.
3Reliability
If fusion parameters are determined through iterative boosting schemes, then authentication performance is improved, but computational complexity increases
Solution Approach 1:
The patent determines fusion parameters through iterative boosting schemes during an offline training phase, using labeled training data to optimize the combination of classifier scores. This preliminary action allows the system to learn optimal fusion parameters in advance, so that during actual authentication operations, the pre-determined parameters can be applied directly without requiring complex iterative computations, thus improving authentication performance while managing online computational complexity.
Data Source
AI summary
A user authentication method includes receiving a first input image including information on a first modality; receiving a second input image including information on a second modality; determining at least one first score by processing the first input image based on at least one first classifier, the at least one first classifier being based on the first modality; determining at least one second score by processing the second input image based on at least one second classifier, the at least one second classifier being based on the second modality; and authenticating a user based on the at least one first score, the at least one second score, a first fusion parameter of the at least one first classifier, and a second fusion parameter of the at least one second classifier.