Meter Audio Signature Detection Through Arrival-Time Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing audio fingerprinting technologies struggle with accurately identifying media in noisy environments due to reliance on loudest audio parts, which often contain background noise, reducing the effectiveness of generated fingerprints and computational efficiency.
Innovation Solution
The method involves removing audio signals based on phase differences between transformed audio signals to enhance fingerprint accuracy and reduce processing burden, utilizing peak values of transformed audio signals for improved signature recognition.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If audio fingerprinting uses the loudest audio parts for identification, then the signal strength is improved, but the accuracy of media identification deteriorates due to background noise contamination
Solution Approach 1:
The patent segments the audio signal into multiple frequency bins and time frames, then selectively processes only those segments that contain relevant media information rather than processing the entire audio signal. This allows the system to focus computational resources on clean signal portions while ignoring noisy segments.
Solution Approach 2:
The patent applies different processing qualities to different parts of the audio signal based on their local characteristics. Frequency bins with high confidence values (indicating clean signal) receive full processing attention, while bins with low confidence values (indicating noise contamination) are downweighted or excluded from fingerprint generation.
2Loss of information
If audio fingerprinting processes the entire audio signal, then the completeness of analysis is improved, but the computational efficiency deteriorates
Solution Approach 1:
The patent extracts only the essential components needed for accurate media identification - specifically, the peak frequency bins with highest confidence values - and excludes redundant or noisy portions from further processing. This extraction approach maintains analysis effectiveness while dramatically reducing computational burden.
Solution Approach 2:
The patent applies partial action by processing only a subset of frequency bins and time frames that contain the most relevant information for identification. Rather than exhaustively analyzing the entire audio signal, the system focuses computational effort on the most informative segments, achieving accurate results with reduced processing.
3Loss of information
If audio fingerprinting includes all frequency components, then the comprehensiveness of signature generation is improved, but the noise interference increases
Solution Approach 1:
The patent dynamically changes the parameters of frequency bin selection based on confidence values calculated from the audio signal characteristics. Frequency bins with confidence values above a threshold are included in the fingerprint, while those below the threshold are excluded. This parameter-based filtering maintains signature comprehensiveness for clean components while eliminating noisy components.
Data Source
AI summary
Methods, apparatus, systems and articles of manufacture are disclosed to improve detection of audio signatures. An example apparatus includes at least one memory, instructions in the apparatus, and processor circuitry to execute the instructions to: determine a first time difference of arrival for a first audio sensor of a meter and a second audio sensor of the meter based on a first audio recording from the first audio sensor and a second audio recording from the second audio sensor; determine a second time difference of arrival for the first audio sensor and a third audio sensor of the meter based on the first audio recording and a third audio recording from the third audio sensor;determine a match by comparing the first time difference of arrival to i) a first virtual source time difference of arrival and ii) a second virtual source time difference of arrival; in response to determining that the first time difference of arrival matches the first virtual source time difference of arrival, identify a first virtual source location as the location of a media presentation device presenting media; and remove the second audio recording to reduce a computational burden on the processor.