Voiceprint Feature Correlation for More Accurate User Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing voiceprint recognition methods rely solely on similarity measures, such as Euclidean distance, leading to low accuracy in identifying whether voiceprint features belong to the same user, as they fail to consider the correlation between voiceprint features.
Innovation Solution
A voiceprint recognition method that combines similarity and correlation analysis to determine user information, utilizing a voiceprint library to identify sets of similar features and calculate correlations based on density coefficients, improving accuracy and robustness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If only similarity measures (e.g., Euclidean distance) are used for voiceprint recognition, then the method is simple to implement, but the recognition accuracy is low
Solution Approach 1:
The patent combines similarity measurement and correlation analysis into a unified voiceprint recognition framework. By merging these two complementary approaches, the system achieves higher recognition accuracy than using either method alone, while maintaining a manageable level of complexity through integrated processing.
Solution Approach 2:
The patent introduces correlation coefficients as an additional parameter alongside traditional similarity measures. By changing the parameter set from only distance metrics to include correlation analysis, the system captures more aspects of voiceprint characteristics, thereby improving recognition accuracy.
2Measurement precision
If correlation analysis between voiceprint features is performed, then the recognition accuracy is improved, but the computational complexity increases
Solution Approach 1:
The patent segments the voiceprint recognition process into distinct stages: similarity measurement, correlation analysis, and integrated decision-making. This segmentation allows computational resources to be allocated efficiently to each stage, reducing overall computational complexity while maintaining high accuracy through targeted analysis at each step.
Solution Approach 2:
The patent applies partial correlation analysis by focusing on specific feature pairs or subsets of voiceprint features rather than computing all possible correlations. This partial action approach maintains improved accuracy for critical comparisons while reducing unnecessary computational overhead from exhaustive analysis.
Data Source
AI summary
A voiceprint recognition method includes: obtaining sets of second voiceprint features by, for each of first voiceprint features, determining voiceprint features in a voiceprint library similar to the first voiceprint feature as a set of second voiceprint features; obtaining first correlations for the first voiceprint features by, for every two of the first voiceprint features, determining a correlation between the two first voiceprint features based on a first set of second voiceprint features corresponding to one first voiceprint feature, a number of second voiceprint features of the first set, a second set of second voiceprint features corresponding to the other first voiceprint feature, and a number of second voiceprint features of the second set, as a first correlation; and determining user information for each of the first voiceprint features based on the first correlations and first similarities between the first voiceprint features.