Audio Speed Change Ratio Detection for Remixed Music Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing music identification techniques fail to accurately determine that two audio samples are the same song when the music is modified by changing tempo or subjected to remixing, leading to false negatives in video sharing platforms.
Innovation Solution
A computing system that determines a speed change ratio between audio samples by extracting audio features, computing a similarity matrix, identifying peak points, and performing multi-line fitting to calculate the speed change ratio, even in the presence of remixing or other modifications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional music identification techniques are used, then the system is simple and fast, but it fails to accurately identify modified music tracks (sped-up or slowed-down versions)
Solution Approach 1:
The patent segments the music identification process into multiple stages: extracting audio features, computing similarity matrix, detecting peak points, fitting peak lines, and calculating speed change ratio. This segmentation allows the system to handle modified music tracks by analyzing specific characteristics at each stage rather than treating the problem as a single complex task.
Solution Approach 2:
The patent introduces a new dimension of analysis by computing similarity matrices across different time scales and identifying peak lines in the frequency-time domain. This dimensional transformation enables the system to detect speed changes by analyzing the slope of peak lines, which would be invisible to traditional single-dimensional music identification methods.
2Measurement precision
If the system performs detailed audio analysis to detect speed changes, then identification accuracy improves, but processing time increases
Solution Approach 1:
The patent extracts only the most relevant audio features (peak points in the similarity matrix) rather than analyzing the entire audio signal in detail. By taking out and focusing on these critical features, the system achieves accurate speed change detection while minimizing processing time associated with analyzing all audio data points.
Solution Approach 2:
The patent performs partial analysis by computing similarity matrices only for extracted audio features rather than the complete audio signal. This partial action approach provides sufficient information for accurate speed change detection without the excessive processing time that would result from analyzing every aspect of the audio data.
3Adaptability or versatility
If the system handles multiple audio modifications (remixing, tempo changes), then versatility improves, but system complexity increases
Solution Approach 1:
The patent creates a universal identification system that handles multiple types of audio modifications (remixing, tempo changes, sped-up, slowed-down versions) through a single unified approach. The similarity matrix and peak line analysis methodology works across different modification types, providing multi-functionality without requiring separate specialized systems for each modification type.
Data Source
AI summary
A computing system including one or more processing devices configured to receive a first audio sample and a second audio sample. The one or more processing devices determine a speed change ratio between the first audio sample and the second audio sample at least in part by extracting first audio features from the first audio sample and second audio features from the second audio sample. Determining the speed change ratio further includes computing a similarity matrix including between the set of first audio features and the set of second audio features. Determining the speed change ratio further includes identifying peak points in the similarity matrix and identifying one or more peak lines. Determining the speed change ratio further includes computing the speed change ratio based at least in part on respective slopes of the one or more peak lines. The one or more processing devices output the speed change ratio.