Mobile Audio Fingerprinting for Individual TV Viewing Attribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems for tracking television viewing history are limited by their inability to provide individual-level audience estimates, often resulting in false positives and missing data due to the lack of return paths on STBs, incomplete demographic information, and shared device scenarios, which restricts market acceptance and accuracy.
Innovation Solution
Implementing a media correlation method using audio detection on mobile devices to identify programs, where smartphones detect audio signals from TVs and sync data with STBs, enabling individual attribution and overcoming limitations by combining data streams from both devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If STB-based television ratings are used to track viewing history, then household-level viewing data can be collected, but individual-level audience estimates cannot be provided due to lack of return paths on STBs and shared device scenarios
Solution Approach 1:
The patent introduces a mobile device as an intermediary between the STB and the rating system. The mobile device captures audio content from the STB, identifies programs through audio fingerprinting, and transmits viewing data to the rating system, thereby bridging the information gap caused by STBs lacking return paths.
Solution Approach 2:
The patent replaces the mechanical/physical limitation of STBs (no return path) with an electronic/digital solution using mobile devices. Instead of modifying the STB hardware, the system uses software-based audio capture and identification on mobile devices to achieve individual-level tracking.
2Quantity of substance
If sampling methods are used to estimate viewing history, then population coverage can be increased, but false positives and missing data occur due to incomplete demographic information
Solution Approach 1:
The mobile device automatically performs audio capture, program identification, and data transmission without requiring user intervention. The system self-services by continuously monitoring audio content and autonomously reporting viewing history, eliminating manual sampling errors and improving data reliability.
Solution Approach 2:
The system establishes a feedback loop where the mobile device continuously monitors audio content, compares it against a database of known programs, and adjusts its reporting based on match results. This feedback mechanism reduces false positives by verifying program identity through audio fingerprinting rather than relying solely on demographic sampling.
3Measurement precision
If audio detection on mobile devices is implemented to identify programs, then individual-level viewing attribution is achieved, but device complexity increases due to combining data streams from multiple devices
Solution Approach 1:
The patent segments the viewing measurement system into distinct functional components: the STB for content delivery, the mobile device for audio capture and identification, and the rating system for data aggregation. Each component performs a specific function, simplifying the overall system architecture despite the multi-device involvement.
Solution Approach 2:
The mobile device performs multiple functions within the system: it acts as an audio recorder, program identifier through fingerprinting, and data transmission device. This multi-functionality reduces the need for separate dedicated components, thereby managing complexity while achieving individual-level attribution.
Data Source
AI summary
A media correlation method, executed by a processor, determines an identity of a program. The method includes detecting a sign-in by a first media device; determining a location of the first media device relative to a second media device; receiving from the first media device, content clips emanating from the second media device; and identifying a program displayed on the second media device based on the received content clips.