Homomorphic Encryption for Biometric Matching Without Decryption
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional biometric systems face limitations in one-to-many searching, security vulnerabilities from faked or replayed biometric signals, and inefficient key management, leading to compromised privacy and authentication accuracy.
Innovation Solution
A privacy-enabled authentication system using deep neural networks (DNNs) processes encrypted biometric feature vectors for one-to-many matching, incorporates liveness checks, and employs homomorphic encryption for secure comparisons, ensuring privacy and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional biometric systems store and compare unencrypted biometric data, then authentication speed is improved, but security and privacy are compromised
Solution Approach 1:
The system segments biometric authentication into two independent components: encrypted biometric templates stored in the database and decryption keys held by users. This segmentation allows the database to store encrypted data without compromising security, while authentication speed is maintained through efficient encrypted search algorithms that operate on the segmented encrypted representations.
Solution Approach 2:
Homomorphic encryption serves as an intermediary mechanism that enables computations on encrypted data without decryption. The system uses this intermediary to perform biometric matching operations directly on encrypted templates, eliminating the need to decrypt sensitive biometric data while maintaining authentication functionality and speed.
2Reliability
If one-to-many biometric searching is implemented on encrypted biometric data, then security is improved, but computational complexity increases
Solution Approach 1:
The system changes the parameter representation of biometric data by transforming raw biometric inputs into encrypted feature vectors through neural networks. This parameter transformation enables efficient one-to-many searching on encrypted data by operating in the encrypted feature space rather than on raw biometric data, reducing computational complexity while maintaining security.
Solution Approach 2:
The patent replaces traditional mechanical biometric matching systems with a neural network-based system that processes encrypted data. The neural networks perform pattern recognition and matching operations directly on encrypted feature vectors, substituting complex cryptographic operations with more efficient neural network computations that maintain security while reducing overall computational complexity.
3Reliability
If liveness detection is added to prevent spoofing, then security against faked biometrics is improved, but system complexity and processing time increase
Solution Approach 1:
The system merges liveness detection functionality directly into the biometric authentication process by using the same neural network infrastructure for both tasks. The neural networks simultaneously perform feature extraction, encryption, and liveness assessment, combining multiple security functions into a unified system that reduces overall complexity while maintaining robust spoofing prevention.
Data Source
Figure 1
Figure 2A
Figure 2B
AI summary
A set of distance measurable encrypted feature vectors can be derived from any biometric data and/or physical or logical user behavioral data, and then using an associated deep neural network ("DNN") on the output (i.e., biometric feature vector and/or behavioral feature vectors, etc.) an authentication system can determine matches or execute searches on encrypted data. Behavioral or biometric encrypted feature vectors can be stored and/or used in conjunction with respective classifications, or in subsequent comparisons without fear of compromising the original data. In various embodiments, the original behavioral and/or biometric data is discarded responsive to generating the encrypted vectors. In another embodiment, distance measurable or homomorphic encryption enables computations and comparisons on cypher-text without decryption of the encrypted feature vectors. Security of such privacy enabled embeddings can be increased by implementing an assurance factor (e.g., liveness) to establish a submitted credential has not been spoofed or faked.