Duplicate Voice Recording Identification via Pitch Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for identifying duplicate voice recordings are inefficient, requiring manual listening and speech-to-text conversion, which are time-consuming and language-dependent, and fail to accurately distinguish between recordings of different lengths.
Innovation Solution
A method that receives digital voice recordings, segments them, extracts pitch values, estimates total voice time, removes insignificant values, identifies unique pitch frequencies, normalizes their occurrence, calculates average pitch and distribution percentiles, and compares these metrics within user-definable thresholds to declare duplicates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual listening and speech-to-text conversion are used to identify duplicate voice recordings, then identification accuracy can be achieved, but processing time increases significantly
Solution Approach 1:
The patent extracts and compares only the essential acoustic feature (pitch values) from voice recordings, rather than analyzing the entire recording content. This extraction approach maintains identification accuracy while dramatically reducing processing time by focusing only on the discriminative feature needed for duplicate detection.
Solution Approach 2:
The patent replaces the mechanical process of manual listening with an automated acoustic feature extraction and comparison system. By substituting human auditory analysis with computational pitch value analysis, the system achieves both time efficiency and consistent identification accuracy.
2Adaptability or versatility
If speech-to-text conversion is used for duplicate identification, then text-based comparison can be performed, but performance becomes highly dependent on language, dialect, and content
Solution Approach 1:
The patent extracts pitch values directly from the acoustic signal, bypassing the speech-to-text conversion process entirely. This extraction of fundamental acoustic features eliminates dependence on language, dialect, and content, making the system universally applicable while maintaining reliable duplicate identification.
Solution Approach 2:
The patent changes the analysis parameter from linguistic content (text) to acoustic physical property (pitch). By comparing pitch values rather than transcribed text, the system achieves language independence while maintaining identification reliability through direct acoustic feature comparison.
3Measurement precision
If traditional duplicate detection methods are used, then all message attributes must be matched, but this increases system complexity
Solution Approach 1:
The patent extracts and compares only the pitch value feature from voice recordings, ignoring other message attributes. This selective extraction approach reduces system complexity by focusing on a single discriminative feature while maintaining duplicate detection accuracy through effective pitch-based comparison.
Data Source
AI summary
A method of identifying duplicate voice recording by receiving digital voice recordings, selecting one of the recordings; segmenting the selected recording, extracting a pitch value per segment, estimating a total time that voice appears in the recording, removing pitch values that are less than and equal to a user-definable value, identifying unique pitch values, determining the frequency of occurrence of the unique pitch values, normalizing the frequencies of occurrence, determining an average pitch value, determining the distribution percentiles of the frequencies of occurrence, returning to the second step if additional recordings are to be processed, otherwise comparing the total voice time, average pitch value, and distribution percentiles for each recording processed, and declaring the recordings duplicates that compared to within a user-definable threshold for total voice time, average pitch value, and distribution percentiles.