Ambient Audio Fingerprinting for Broadcast Viewership Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Analyzing viewership of broadcast content, especially for online and Over-the-Air (OTA) content, is challenging due to the difficulty in detecting whether a user has actually viewed the content and for how long, particularly in environments where traditional tracking methods are ineffective.
Innovation Solution
A method that records ambient audio in a household to generate an ambient audio fingerprint, which is used to determine if a user has viewed broadcast content by correlating the audio features with the content item, allowing for accurate impression logging and user profile updates within an online system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional tracking methods are used to monitor broadcast content viewership, then the system is simple to implement, but the measurement precision is insufficient to accurately detect whether a user has actually viewed the content and for how long
Solution Approach 1:
The patent introduces ambient audio recording as an intermediary medium to indirectly detect content viewership. Instead of directly tracking user viewing behavior, the system records ambient audio in the user's environment and analyzes it for content-specific audio features, watermarks, or identifiers. This intermediary approach enables precise viewership measurement without requiring direct interaction with or complex tracking of the user's viewing device.
Solution Approach 2:
The patent replaces traditional mechanical/electronic tracking systems (such as cookies, device identifiers, or direct sensor-based tracking) with an acoustic field-based detection system. By substituting the mechanical tracking infrastructure with audio recording and analysis, the system achieves precise viewership detection while avoiding the complexity of direct user device tracking infrastructure.
2Measurement precision
If ambient audio recording is used to accurately detect content viewership, then the measurement precision is improved, but the loss of information increases due to the difficulty of correlating audio features with specific content items
Solution Approach 1:
The patent transforms the audio signal parameters by extracting specific features (such as frequency spectra, temporal patterns, or acoustic fingerprints) from the ambient audio recording. By changing the parameter representation from raw audio waveforms to extracted acoustic features, the system enables reliable correlation with content identifiers while reducing information loss and improving content identification accuracy.
Solution Approach 2:
The patent creates an acoustic copy or fingerprint of the content's audio signature that can be independently verified. By extracting and storing reference audio features from the broadcast content, the system can compare these against ambient audio recordings to confirm viewership without requiring the original content file or direct connection to the broadcasting system, thus preserving information integrity.
3Measurement precision
If the client device continuously records ambient audio to detect content viewing, then the measurement precision is improved, but the use of energy increases
Solution Approach 1:
The patent implements periodic or event-triggered audio recording instead of continuous recording. The system activates the audio recording function only during specific periods (such as when content playback is detected or during scheduled intervals) or when specific events occur (such as detecting a watermark or identifier). This periodic operation maintains measurement precision while dramatically reducing energy consumption compared to continuous recording.
Solution Approach 2:
The patent performs preliminary detection of content playback conditions (such as detecting broadcast signals, watermarks, or identifiers in the audio stream) before activating the ambient audio recording. By preliminarily identifying when content is being played, the system avoids unnecessary continuous recording and only activates energy-intensive audio capture when actually needed for viewership detection, thus optimizing energy usage.
Data Source
AI summary
An online system analyzes broadcast content viewed by individuals in a household. Each individual in the household is associated with a client device on which a software application module is executed. When the software application module detects one or more broadcasting signals of a content item broadcasted to the household, the software application module records the ambient audio, including audio from the broadcasting device. The software application module sends an identifier of the individual associated with the client device, an ambient audio fingerprint derived from the recorded ambient audio, and time information for the recorded ambient audio to the online system. The online system, based on the ambient audio data, identifies the corresponding individual and content item and logs an impression for the content item upon determination that there was an impression of the identified content item by the identified individual.