Streaming Session Identification via Watermark Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for determining audience measurement data for streaming sessions face challenges in accurately associating panelists with specific viewing sessions, especially in the context of streaming services where traditional television audience measurement techniques are not effective.
Innovation Solution
The proposed solution involves the use of identifiers, such as watermarks and fingerprints, embedded in streaming content. These identifiers, when detected during streaming sessions, are logged along with the time of detection. This data is then used to identify panelists and associate them with specific streaming sessions through comparison with streaming session logs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional television audience measurement techniques are used for streaming sessions, then the measurement process remains simple and familiar, but the accuracy of associating panelists with specific viewing sessions deteriorates
Solution Approach 1:
The patent introduces watermarks and fingerprints as intermediary elements embedded in streaming content. These intermediaries serve as unique identifiers that bridge the gap between panelist devices and specific streaming sessions, enabling accurate association without requiring direct identification methods. The watermark detector and fingerprint analyzer capture these intermediaries to link viewing sessions with panelist identifiers.
Solution Approach 2:
The patent creates copies of unique identifiers (watermarks and fingerprints) that are embedded within streaming content. These copied identifiers are then detected and matched against session logs to establish connections between panelists and viewing sessions. The system captures copies of these identifiers at multiple points in the streaming chain to ensure accurate tracking.
2Measurement precision
If identifiers such as watermarks and fingerprints are embedded in streaming content, then the accuracy of session identification improves, but the complexity of the measurement system increases
Solution Approach 1:
The patent divides the identification system into separate functional components: watermark embedding, fingerprint generation, detection mechanisms, and matching algorithms. Each component handles a specific aspect of identifier processing, allowing the system to manage complexity through modular design while maintaining high identification accuracy across multiple streaming sessions.
3Reliability
If multiple identifiers are used to track streaming sessions, then the reliability of audience measurement data improves, but the loss of information due to processing complexity increases
Solution Approach 1:
The patent implements feedback mechanisms where detected watermarks and fingerprints are continuously matched against session logs, and results are used to refine and confirm panelist associations. This feedback loop ensures that multiple identifiers work together synergistically to improve reliability while minimizing information loss through iterative verification and cross-validation of measurement data.
Data Source
AI summary
Methods and apparatus to identify streaming sessions are disclosed. A disclosed apparatus for identifying streaming sessions includes an encoder to generate an identifier corresponding to a media file of an advertisement, a detector to determine a presence of the identifier in the advertisement presented at a media presentation device, a timing analyzer to determine a time at which the advertisement was presented, and an associator. The associator is to compare the identifier and the time to a streaming session log from a streaming service provider, identify a streaming session in which the advertisement associated with the identifier was presented at the time based on the comparison, and associate a panelist identifier corresponding to the media presentation device with a streaming session identifier corresponding to the streaming session.