Voice Biometric Passphrase Optimization via Phonetic Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Text-independent voice biometric systems experience lower accuracy in active authentication scenarios, particularly with short segments of speech, compared to text-dependent systems, due to the lack of lexical match during enrollment and authentication.
Innovation Solution
A method that generates a set of candidate passphrases, calculates their phonetic content, and selects a subset that matches a user's text-independent voiceprint, allowing for accurate authentication without explicit enrollment and enabling passphrase changes, by employing phonetic content analysis and Universal Background Model state occupancy vectors to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If text-independent voice biometric model is used to skip enrollment stage, then ease of operation is improved, but measurement precision deteriorates
Solution Approach 1:
The system performs preliminary phonetic analysis of candidate passphrases and compares them against the voiceprint model before authentication. This preliminary matching action ensures that only phonetically compatible passphrases are presented to the user, thereby maintaining high authentication accuracy even without traditional enrollment
Solution Approach 2:
The system changes the parameter of passphrase selection from random or static to dynamically optimized based on phonetic content analysis. By adjusting the passphrase selection criterion to match the user's voice characteristics, the system maintains measurement precision while using text-independent modeling
2Measurement precision
If text-dependent voice biometric system is used with predefined text, then measurement precision is improved, but adaptability deteriorates
Solution Approach 1:
The system applies local quality by maintaining high phonetic matching requirements specifically for the passphrase portion while allowing flexibility in other aspects of the authentication system. Each candidate passphrase is evaluated individually against the user's voiceprint, ensuring local optimization of authentication accuracy without requiring global text dependency
Solution Approach 2:
The system implements dynamic passphrase selection where the optimal passphrase is chosen based on real-time phonetic analysis of the user's voice characteristics. This dynamic adaptation allows the system to maintain high accuracy while being versatile across different users and speech patterns, unlike static text-dependent systems
Data Source
AI summary
Systems and methods for optimizing matched voice biometric passphrases may generate a set of k candidate passphrases; calculate a phonetic content of each candidate passphrase of the set of k candidate passphrases; at least one of during and after a voice biometric enrollment of a user, calculate a user phonetic content of a text-independent voiceprint of the user, wherein the text-independent voiceprint of the user was captured during the voice biometric enrollment; identify a subset of j passphrases from the set of k candidate passphrases, wherein each of the subset of j passphrases meets a match threshold in phonetic content with the voiceprint of the user; one of prior to and during a user authentication, select a first one of the subset of j passphrases; and present the selected first one of the subset of j passphrases to the user for use in the user authentication.