Homomorphic Encryption Biometric Database Comparison via Clustering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current biometric database comparison methods, especially using convolutional neural networks (CNNs), are computationally burdensome and insecure, particularly when trying to determine if an individual is present in multiple confidential databases, as they require extensive comparisons of encrypted data.
Innovation Solution
A method involving the use of homomorphic encryption and a classification model to group similar biometric data into bins, allowing for efficient comparisons between encrypted databases, reducing the number of necessary comparisons by grouping elements, and ensuring confidentiality through encrypted domain operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If homomorphic encryption is used to compare confidential biometric databases, then data security is maintained, but computational burden increases significantly
Solution Approach 1:
The patent divides the biometric database into multiple clusters based on similarity metrics. Instead of comparing every element across databases, the system groups elements into clusters and performs comparisons at the cluster level, significantly reducing the number of individual comparisons needed while maintaining security through encrypted operations.
Solution Approach 2:
The patent performs preliminary clustering of biometric data before the actual comparison process. By pre-organizing data into similarity-based clusters and encrypting them in advance, the system reduces the computational complexity of the subsequent comparison operations, allowing efficient searching without compromising security.
2Reliability
If all elements in one database are compared with all elements in another database, then complete matching is achieved, but the number of comparisons becomes extremely large
Solution Approach 1:
The patent segments the database elements into clusters based on biometric similarity. By organizing elements into groups with comparable characteristics, the system reduces the comparison space from O(n×m) to a much smaller number of cluster-level comparisons, while still ensuring that potential matches are not missed.
Solution Approach 2:
The patent applies different comparison strategies to different clusters based on their characteristics. High-density clusters may receive more thorough examination while low-density clusters use lighter comparison methods, optimizing the balance between matching completeness and computational efficiency.
Data Source
AI summary
A method for comparing a first and a second databases to determine whether an individual is represented by both an element of the first database and an element of the second database, wherein said elements are biometric data, including the implementation of the steps applying a classification model for each element of each database so as to construct a set of first and second bins of the respective first and second databases, each bin bringing together similar elements, each first bin being associated with a second bin; comparing the elements from the first database for at least one pair of an associated first bin and second bin belonging to said first bin with the elements from the second database belonging to said second bin, at least one of the first and the second databases then being encrypted homomorphically.