Voice Biometric Authentication With Multimodal Identity Filtering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voice biometric systems require separate enrollment with each vendor system, limiting user experience and adoption across diverse devices and public settings, and face challenges in resource-intensive identification processes and false acceptance/rejection rates.
Innovation Solution
A centralized machine-learning architecture enables seamless voice biometrics across vendors by generating and managing a single point of enrollment, using multi-modal data fusion to reduce resource demands and false acceptance/rejection rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If separate enrollment with each vendor system is required, then each vendor can maintain independent control over their voice biometric data, but user experience deteriorates and adoption rates decrease due to the need to register with multiple systems
Solution Approach 1:
The patent introduces a centralized voice biometric verification service that acts as an intermediary between users and multiple vendor systems. Instead of requiring users to enroll separately with each vendor, the service provides a single enrollment point that issues verification tokens valid across multiple vendors, eliminating the need for repeated enrollments while maintaining vendor data control
Solution Approach 2:
The verification service is designed to be universal, serving multiple vendor systems through a single interface. The service maintains a database of verified users and issues verification tokens that can be presented to any enrolled vendor, making the system multi-functional and vendor-agnostic while simplifying the user experience
2Reliability
If voice biometric systems are deployed privately only, then security and accuracy are maintained, but expansion to public settings and multi-user scenarios is limited
Solution Approach 1:
The system dynamically adapts its verification behavior based on the deployment context. In private settings, it maintains high-security verification protocols, while in public settings it enables multi-user scenarios and expanded access. The service can adjust verification strictness and user management based on the specific use case, allowing seamless transition between private and public deployments
3Ease of operation
If a centralized verification service is implemented, then user enrollment is simplified and cross-vendor access is enabled, but system complexity and infrastructure requirements increase
Solution Approach 1:
Instead of requiring each vendor to maintain their own complete voice biometric verification system, the patent uses a centralized service that copies and distributes verification authority. The centralized service creates a master verification database and issues tokens that can be presented to any vendor, eliminating the need for redundant infrastructure at each vendor while simplifying the overall system architecture
4Adaptability or versatility
If voice biometric systems are used across multiple vendors, then user accessibility is improved, but false acceptance and false rejection rates increase due to varying system characteristics
Solution Approach 1:
The system changes the verification parameters and matching thresholds dynamically based on the specific vendor system being accessed. The centralized service adjusts verification strictness, feature extraction methods, and matching algorithms to optimize for each vendor's specific characteristics, thereby maintaining low false acceptance and rejection rates across diverse vendor systems while preserving cross-vendor compatibility
Data Source
AI summary
Disclosed are systems and methods including computing-processes executing machine-learning architectures extract vectors representing disparate types of data and output predicted identities of users accessing computing services, without express identity assertions, and across multiple computing services, analyzing data from multiple modalities, for various user devices, and agnostic to architectures hosting the disparate computing service. The system invokes the identification operations of the machine-learning architecture, which extracts biometric embeddings from biometric data and context embeddings representing all or most of the types of metadata features analyzed by the system. The context embeddings help identify a subset of potentially matching identities of possible users, which limits the number of biometric-prints the system compares against an inbound biometric embedding for authentication. The types of extracted features originate from multiple modalities, including metadata from data communications, audio signals, and images. In this way, the embodiments apply a multi-modality machine-learning architecture.