Dynamic Video Insertion Point Detection via Scene Analysis

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Traditional content distribution techniques struggle with dynamic insertion of supplemental content, such as advertisements, in video streaming, especially in live or near-live scenarios, as they rely on manual marker insertion which is not suitable for real-time adjustments.

Innovation Solution

A video packaging and origination service that automatically determines insertion points for supplemental content by analyzing scene changes, activity levels, and third-party data, and selects relevant content based on image recognition, textual analysis, and user preferences, enabling dynamic content insertion and selection.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If manual marker insertion is used for content insertion, then content placement is simple to implement, but real-time adjustment capability is lost

Engineering Contradiction:

Improvecontent insertion implementationVSAvoidreal-time adjustment capability

Core Design Contradiction:

Ease of operationVSAdaptability or versatility

Solution Approach 1:

The system automatically determines insertion points by analyzing video content characteristics such as scene changes and keyframes, eliminating the need for manual marker insertion while enabling real-time adaptation to different video streams and content types

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The insertion point determination is made dynamically during video streaming based on real-time analysis of video data characteristics, allowing the system to adapt to live and on-demand content without pre-defined markers

Inventive Principle:

Principle #15Dynamics

2Adaptability or versatility

If dynamic content insertion is implemented, then user engagement and ad relevance improve, but system complexity increases

Engineering Contradiction:

Improvedynamic content insertion capabilityVSAvoidsystem complexity

Core Design Contradiction:

Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The system identifies and marks potential insertion points during the video encoding and packaging phase, before actual content delivery. This preliminary analysis creates a framework that simplifies real-time content insertion during streaming by pre-determining optimal locations based on scene changes and video structure

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system introduces an intermediary processing layer between video encoding and content delivery that automatically analyzes video characteristics and determines insertion points. This intermediary component handles the complexity of dynamic analysis, allowing the rest of the system to operate with simpler logic

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS11064237B1Automatically generating content for dynamically determined insertion points

Publication Date: 2021.07.13 AMAZON TECH INC

AI summary

A video packaging and origination service can process requests for content segments from requesting user devices. The video packaging and origination service can processing video attributes, audio attributes and social media feeds to dynamically determine insertion points for supplemental content. Additionally, the video packaging and origination service can identify supplemental content utilizing the same attribute information.