Streaming Content Context Analysis for Real-Time Ad Insertion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
OTT media services face challenges in delivering appropriate additional content during streaming due to incomplete information about the current content being viewed, as traditional mechanisms lack the ability to dynamically provide content that aligns with the show being watched, especially in environments like Free Ad Supported Television (FAST) services.
Innovation Solution
Embodiments leverage incomplete information in a request for additional content by combining it with publicly available data, such as Electronic Programming Guides (EPGs), to identify the specific program being viewed, and utilize analysis modules to determine appropriate content for insertion, including techniques like object recognition, speech-to-text, and natural language processing to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional content delivery mechanisms are used in OTT streaming, then content can be delivered to users, but the system cannot dynamically identify the current program being viewed to deliver appropriate additional content
Solution Approach 1:
The system implements feedback by analyzing the live video stream content itself to identify the current program being viewed. Analysis modules process the video content and provide information about what is being displayed, which then feeds back into the content delivery system to determine appropriate additional content to insert during stream breaks.
Solution Approach 2:
Analysis modules serve as intermediaries between the live video stream and the content delivery system. These modules extract contextual information from the video content and translate it into usable data that informs subsequent content delivery decisions, bridging the gap between raw video and intelligent content selection.
2Measurement precision
If the system downloads and analyzes the entire live stream to identify contextual information, then accurate program identification is achieved, but significant processing time and computational resources are required
Solution Approach 1:
The system performs preliminary actions by downloading and analyzing the live video stream in advance, before additional content needs to be inserted. This allows the system to identify the current program and prepare contextual information ahead of time, so that when a stream break occurs, appropriate content can be delivered immediately without delay.
Solution Approach 2:
The system analyzes more video content than strictly necessary by examining the entire live stream rather than just key frames or metadata. This excessive analysis ensures complete and accurate program identification, capturing all contextual information that might be relevant for subsequent content delivery decisions.
3Loss of information
If analysis modules process video content to identify contextual information, then relevant content can be identified for additional content delivery, but the system complexity increases
Solution Approach 1:
The analysis modules are designed with multi-functionality to handle various types of content analysis. A single modular system can identify different types of programs (sports, news, entertainment, etc.) and extract diverse contextual information, reducing the need for separate specialized systems for each content type and thereby managing complexity through universal design.
Data Source
AI summary
Systems and methods for identifying contextual information in streams of content being delivered by streaming platforms are discussed. The identified contextual information may be used to insert additional content in the stream of content. The identification of contextual information and the insertion of additional content based upon the identification may be performed in near real-time during delivery of the stream.