Audio Fingerprinting Using Energy Regions for Timing Detection

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional audio content recognition systems lack the ability to detect the timing position within audio content, are hindered by large fingerprint sizes, and result in complex and delayed recognition processes.

Innovation Solution

The development of audio content recognition algorithms that generate compact fingerprints by analyzing audio energy characteristics, allowing for real-time detection of content identity and timing position, using techniques such as energy region identification, hashing, and MinHash to reduce data complexity and enhance search efficiency.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional audio fingerprinting algorithms are used, then audio content recognition is achieved, but the fingerprint size becomes large and processing complexity increases

Engineering Contradiction:
Improveaudio content recognition accuracyVSAvoidfingerprint size and processing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent extracts only the essential energy characteristics from audio signals to create fingerprints. Instead of using complete spectral information, it selectively captures energy values at specific time-frequency points, removing redundant data while preserving recognition capability. This extraction approach reduces fingerprint size and processing complexity while maintaining recognition accuracy.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent creates simplified copies of audio content in the form of energy-based fingerprints rather than using the original complex audio data. These fingerprint copies contain condensed information about audio energy patterns, enabling efficient comparison and recognition without requiring processing of the full audio signal, thus reducing computational complexity and data size.

Inventive Principle:
Principle #26Copying

2Measurement precision

If conventional audio fingerprinting algorithms are used, then audio content identity is identified, but timing position detection within the content is not achieved

Engineering Contradiction:
Improvecontent identity recognitionVSAvoidtiming position information
Core Design Contradiction:
Measurement precisionVSLoss of information

Solution Approach 1:

The patent segments the audio signal analysis into distinct time-based energy measurements. By dividing the audio stream into sequential analysis windows and tracking energy characteristics at each segment, the system can identify both the content identity and the specific timing position where the content occurs. This segmentation preserves temporal information that would otherwise be lost in conventional fingerprinting.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent adds a temporal dimension to the fingerprinting process by incorporating time-position information into the energy-based fingerprint structure. Instead of creating static fingerprints, the system generates time-stamped energy measurements that exist in both frequency-energy space and time space, enabling simultaneous identification of content identity and timing position through multi-dimensional fingerprint comparison.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Measurement precision

If conventional audio recognition processes are used, then content identification is performed, but recognition delay is increased

Engineering Contradiction:
Improvecontent recognition accuracyVSAvoidrecognition delay
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent applies partial action by performing recognition on selected energy characteristics rather than analyzing the complete audio signal. It focuses computational resources on measuring energy at critical time-frequency points identified through the analysis window, rather than processing all audio data uniformly. This selective partial processing reduces computation time and recognition delay while maintaining sufficient accuracy for content identification.

Inventive Principle:
Principle #16Partial or excessive action

Solution Approach 2:

The patent implements preliminary action by pre-defining analysis windows and energy measurement points before actual recognition occurs. The system prepares the fingerprint extraction framework in advance, establishing the temporal and spectral parameters for energy measurement, which enables faster real-time recognition without requiring complex on-the-fly parameter optimization, thus reducing overall recognition delay.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3440564B1Audio fingerprinting based on audio energy characteristics
Publication Date: 2021.01.13 SOURCE DIGITAL INC
  • EP3440564B1 patent drawingFigure 1
  • EP3440564B1 patent drawingFigure 2
  • EP3440564B1 patent drawingFigure 2A

AI summary

Audio fingerprinting includes obtaining audio samples of a piece of audio, generating frequency representations of the audio samples, identifying increasing and decreasing energy regions in frequency bands of the frequency representations, and generating hashes of features of the piece of audio. Each hash of features corresponds to portions of the identified energy regions appearing in a respective time window. Each feature is defined as a numeric value that encodes information representing: a frequency band of an energy region appearing in the respective time window, whether the energy region appearing in the respective time window is an increasing energy region or whether the energy region appearing in the respective time window is a decreasing energy region, and a placement of the energy region appearing in the respective time window.