Audio Fingerprinting for Real-Time Advertisement Detection

Solution Overview

Problem

Existing audio fingerprinting techniques are slow and resource-intensive, making them unsuitable for real-time identification of advertisements or segments in audio streams during streaming services, as they require significant processing power and storage capacity.

Innovation Solution

A fast and less resource-intensive method using a compact audio fingerprint generated from a spectrogram of the audio stream, which is processed to identify characteristics such as advertisements, allowing for quick identification and control of content delivery.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Speed

If traditional audio fingerprinting techniques are used, then identification accuracy is maintained, but processing speed decreases and resource consumption increases

Engineering Contradiction:

Improveprocessing speedVSAvoidresource consumption

Core Design Contradiction:

SpeedVSDevice complexity

Solution Approach 1:

The patent extracts only the essential features from the audio signal by using spectrogram analysis to identify peak frequency bins, rather than processing the entire audio fingerprint data. This extraction approach maintains identification accuracy while significantly reducing processing speed requirements and resource consumption.

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The audio fingerprint matching process is segmented into distinct phases: spectrogram generation, peak detection, and hash comparison. By dividing the process into manageable segments, the system can process only critical portions of the audio data, improving speed while reducing overall resource consumption.

Inventive Principle:

Principle #1Segmentation

2Reliability

If traditional audio fingerprinting techniques are used, then identification accuracy is maintained, but storage requirements increase

Engineering Contradiction:

Improveidentification accuracyVSAvoidstorage capacity

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The patent extracts only the essential features from the audio signal by using spectrogram analysis to identify peak frequency bins, rather than processing the entire audio fingerprint data. This extraction approach maintains identification accuracy while significantly reducing processing speed requirements and resource consumption.

Inventive Principle:

Principle #2Taking out (Extraction)

3Adaptability or versatility

If real-time advertisement identification is implemented, then content control capability improves, but processing load increases

Engineering Contradiction:

Improvecontent control capabilityVSAvoidprocessing load

Core Design Contradiction:

Adaptability or versatilityVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary spectrogram analysis and peak detection on audio segments before full fingerprint matching is required. This preliminary action prepares the data in advance, enabling real-time content control decisions to be made with reduced processing load during actual advertisement identification.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The audio fingerprint matching process is segmented into distinct phases: spectrogram generation, peak detection, and hash comparison. By dividing the process into manageable segments, the system can process only critical portions of the audio data, improving speed while reducing overall resource consumption.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS9159328B1Audio fingerprinting for advertisement detection

Publication Date: 2015.10.13 VERIZON PATENT & LICENSING INC

AI summary

A device may receive an audio sample, and may separate the audio sample into multiple sub-band signals in multiple frequency bands. The device may modify an upper boundary and a lower boundary of at least one of the frequency bands to form modified frequency bands. The device may modify the sub-band signals to form banded signals associated with the modified frequency bands. The device may smooth the banded signals to form smoothed signal values. The device may identify peak values included in the smoothed signal values, and may generate an audio fingerprint for the audio sample based on the smoothed signal values and the peak values.