Encrypted Biometric Feature Vectors for Secure One-to-Many Matching
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 inefficiencies in managing encrypted biometric data, leading to compromised security and accuracy in identity validation.
Innovation Solution
A privacy-enabled biometric system utilizing encrypted feature vectors and deep neural networks (DNNs) for secure, scalable one-to-many matching, incorporating liveness detection and randomized biometric checks to authenticate identities, ensuring secure and accurate authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional biometric systems perform one-to-many searching, then identity validation capability is improved, but security is worsened due to vulnerability to faked or replayed biometric signals
Solution Approach 1:
The patent introduces an intermediary verification mechanism that validates the authenticity of biometric signals before processing. This intermediary layer checks for liveness and prevents replay attacks, allowing the system to perform one-to-many searching while maintaining security. The intermediary acts as a gatekeeper between the biometric input and the matching algorithm.
2Reliability
If biometric data is encrypted for security, then security is improved, but searching and matching operations become more complex and less efficient
Solution Approach 1:
The patent performs preliminary encryption of biometric data during the enrollment phase, converting raw biometric templates into encrypted forms. This preliminary action allows subsequent searching and matching operations to be performed on the already-encrypted data without requiring decryption, thus maintaining security while reducing operational complexity during authentication.
3Reliability
If liveness detection and randomized biometric checks are implemented, then security against spoofing is improved, but processing time and system complexity increase
Solution Approach 1:
The patent implements partial liveness detection by applying verification checks selectively rather than to all biometric inputs uniformly. Randomized biometric checks are performed on a subset of samples or at specific intervals during the authentication process. This partial application reduces the overall processing time while still maintaining effective security against spoofing attacks.
Data Source
AI summary
In one embodiment, a set of feature vectors can be derived from any biometric data, and then using a deep neural network (“DNN”) on those one-way homomorphic encryptions (i.e., each biometrics' feature vector) an authentication system can determine matches or execute searches on encrypted data. Each biometrics' feature vector can then be stored and/or used in conjunction with respective classifications, for use in subsequent comparisons without fear of compromising the original biometric data. In various embodiments, the original biometric data is discarded responsive to generating the encrypted values. In another embodiment, the homomorphic encryption enables computations and comparisons on cypher text without decryption of the encrypted feature vectors. Security of such privacy enable biometrics can be increased by implementing an assurance factor (e.g., liveness) to establish a submitted biometric has not been spoofed or faked.