Phoneme Comparison Scoring for Audio Content Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques for audio data recognition and identification are inefficient, inaccurate, and time-intensive, particularly for large datasets, and struggle to determine if an audio stream is a copy of another, requiring substantial manual review.
Innovation Solution
A phoneme comparison and scoring system that detects sequences of phonemes in audio data and compares them against known sequences using mathematical measurements like Euclidean distance, allowing for efficient identification and organization of similar content without the need for phonetic recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional phonetic recognition techniques are used to process audio data, then identification accuracy can be achieved, but processor and memory resources are substantially consumed and processing speed becomes slow
Solution Approach 1:
The patent extracts only the essential phoneme sequence information from audio data, rather than performing complete phonetic recognition. By taking out only the necessary phoneme identifiers and comparing these extracted sequences, the system achieves accurate identification without the substantial computational overhead of full phonetic recognition, thus resolving the contradiction between accuracy and processing speed
Solution Approach 2:
The patent segments audio data into discrete phoneme units and compares only these segmented phoneme sequences. This segmentation approach breaks down the complex task of full audio recognition into manageable phoneme-level comparisons, maintaining identification accuracy while significantly reducing processor and memory resource consumption, thereby improving processing speed
2Reliability
If conventional phonetic recognition techniques are used to determine if audio streams are copies, then identification can be achieved, but substantial manual review and time-intensive processing are required
Solution Approach 1:
The patent implements a self-service system where phoneme sequences are automatically extracted, compared, and scored without requiring manual review. The system autonomously determines whether audio streams are copies by comparing phoneme sequences and applying scoring thresholds, thereby achieving reliable copy identification while eliminating time-intensive manual review processes
Solution Approach 2:
The patent replaces the mechanical manual review process with an automated computational system that extracts phoneme sequences, compares them using mathematical measurements, and generates similarity scores. This substitution of manual mechanical review with automated phoneme-based comparison maintains reliable copy identification while dramatically reducing the time loss associated with manual processing
3Loss of information
If conventional techniques are used to locate copies of data throughout the Internet, then data can be identified, but the techniques do not scale very well and are problematic
Solution Approach 1:
The patent changes the fundamental parameter used for data copy identification from complete audio or text comparison to phoneme sequence comparison. This parameter change enables the system to scale effectively across the Internet because phoneme sequences are compact, can be processed efficiently, and maintain sufficient discriminatory power for identifying data copies, thereby resolving the scalability problem while preserving data copy identification capability
Data Source
AI summary
Content matching using phoneme comparison and scoring is described, including extracting phonemes from a file, comparing the phonemes to other phonemes, associating a first score with the phonemes based on a probability of the other phonemes matching the phonemes, and providing the file with another file when a request is received to access one or more files having a second score that is substantially similar to the first score.