Media Fingerprint Adjustment for Pitch, Time, and Resampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional fingerprinting techniques struggle to identify media that has been pitch shifted, time shifted, or resampled, as they rely on adjusting audio samples to match reference samples, which is inefficient and prone to errors.
Innovation Solution
Generate a sample fingerprint and adjust it to accommodate for pitch shifts, time shifts, and resampling ratios, rather than adjusting the audio sample to match reference fingerprints, allowing for more robust media identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional fingerprinting techniques adjust audio samples to match reference samples, then media identification can be performed, but the process becomes inefficient and prone to errors when dealing with pitch shifted, time shifted, or resampled audio
Solution Approach 1:
Instead of adjusting the audio sample to match the reference fingerprint, the patent inverts the approach by adjusting the reference fingerprint to match the characteristics of the query audio. This involves generating multiple adjusted reference fingerprints with different pitch shifts, time shifts, and resampling ratios, then comparing these adjusted references against the query fingerprint. This inversion resolves the contradiction by making the system robust to various audio transformations while maintaining efficient comparison processes.
Solution Approach 2:
The patent applies parameter changes by systematically varying key parameters of the reference fingerprint including pitch shift values, time shift values, and resampling ratios. By generating adjusted reference fingerprints across a range of these parameters, the system can accommodate pitch shifted, time shifted, and resampled audio without requiring complex sample adjustments, thereby improving both reliability and productivity.
2Adaptability or versatility
If audio samples are adjusted to match reference fingerprints, then identification can proceed, but the complexity of handling various pitch shifts, time shifts, and resampling ratios increases
Solution Approach 1:
The patent simplifies the process by inverting the adjustment approach: instead of adjusting query audio samples to match references, it adjusts reference fingerprints to match various query characteristics. This reduces the complexity of handling transformed audio while maintaining high adaptability, as the system now only needs to generate adjusted references using standardized parameters rather than performing complex sample manipulations.
Solution Approach 2:
The patent applies preliminary action by pre-generating adjusted reference fingerprints with various pitch shifts, time shifts, and resampling ratios before the actual identification process. This preparation step allows the system to quickly compare pre-adjusted references against query fingerprints without performing complex adjustments during real-time identification, thereby reducing process complexity while maintaining versatility.
3Reliability
If multiple adjustments are made to accommodate different audio transformations, then detection capability improves, but the time and computational resources required increase
Solution Approach 1:
The patent efficiently handles multiple adjustments by systematically varying parameters (pitch shift, time shift, resampling ratios) to generate adjusted reference fingerprints. By organizing these adjustments around key parameters rather than performing exhaustive audio sample manipulations, the system maintains high detection capability while reducing processing time and computational resource requirements.
Solution Approach 2:
By pre-generating adjusted reference fingerprints with various parameter transformations before the identification process, the system avoids performing multiple adjustments during real-time query processing. This preliminary preparation maintains comprehensive detection capability while significantly reducing the time and computational resources required during actual media identification operations.
Data Source
AI summary
Methods, apparatus, systems and articles of manufacture are disclosed to identify media. An example method includes: in response to a query, generating an adjusted sample media fingerprint by applying an adjustment to a sample media fingerprint; comparing the adjusted sample media fingerprint to a reference media fingerprint; and in response to the adjusted sample media fingerprint matching the reference media fingerprint, transmitting information associated with the reference media fingerprint and the adjustment.