Private Identity System Using Encrypted Biometric Feature Vectors
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current identification and authentication methods in computing environments are insecure, with user identifiers and passwords being inadequate, and existing augmentations like multifactor authentication failing to fully address security concerns.
Innovation Solution
A private identity system using fully encrypted biometric and behavioral information to securely identify users, enabling seamless functionality across devices and switching between users with minimal overhead, employing pre-trained embedding networks and classification networks to generate and recognize encrypted feature vectors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional user identifiers and passwords are used for authentication, then the system is easy to operate, but the security is insufficient
Solution Approach 1:
The patent replaces traditional mechanical authentication systems (usernames and passwords) with a biometric-based authentication system using neural networks. The system captures biometric data (such as facial images) and processes it through embedding networks and classification networks to verify user identity, thereby improving security while maintaining ease of operation through automatic biometric recognition.
2Reliability
If multifactor authentication is implemented to enhance security, then the security profile improves, but the system complexity and overhead increase
Solution Approach 1:
The patent extracts the essential authentication function from complex multifactor authentication systems and implements a streamlined biometric-based approach. By focusing on biometric data processing through specialized neural networks (embedding networks for feature extraction and classification networks for verification), the system achieves strong security without the overhead of managing multiple authentication factors.
3Measurement precision
If biometric data is processed to improve identification accuracy, then the measurement precision improves, but the computational complexity and processing time increase
Solution Approach 1:
The patent segments the biometric processing task into two distinct neural network components: embedding networks that extract and encode biometric features into compact representations, and classification networks that perform the actual verification. This segmentation allows for optimized processing where the embedding network handles the complex feature extraction once during enrollment and authentication, while the classification network performs faster comparisons, thereby improving identification accuracy while managing computational complexity.
4Reliability
If encrypted feature vectors are used to protect user privacy, then the security and privacy protection improve, but the processing and matching operations become more complex
Solution Approach 1:
The patent introduces encrypted feature vectors as an intermediary representation between raw biometric data and user identification. The embedding networks transform biometric data into encrypted feature vectors that preserve the essential characteristics needed for matching while removing personally identifiable information. The classification networks then operate on these encrypted vectors to perform verification, thereby protecting user privacy while enabling accurate authentication through the intermediary encrypted representation.
Data Source
AI summary
In various embodiments, a fully encrypted private identity based on biometric and/or behavior information can be used to securely identify any user efficiently. According to various aspects, once identification is secure and computationally efficient, the secure identity/identifier can be used across any number of devices to identify a user an enable functionality on any device based on the underlying identity, and even switch between identified users seamlessly all with little overhead. In some embodiments, devices can be configured to operate with function sets that transition seamlessly between the identified users, even, for example, as they pass a single mobile device back and forth. According to some embodiments, identification can extend beyond the current user of any device, into identification of actors responsible for activity/content on the device.