Video Stream Object Recognition for Real-Time User-Specific Information

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Manually searching for information about objects in live and recorded video streams is time-consuming and requires prior knowledge, diverting user attention and failing to provide real-time, user-specific information services.

Innovation Solution

Leveraging machine learning and artificial intelligence to identify objects in video streams and provide real-time, user-specific information services by interacting with objects directly within the stream, using AI models on client or server sides, and delivering metadata through MPEG-7 streams.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Loss of information

If manual searching is performed to find information about objects in video streams, then information can be obtained, but user attention is diverted and time is lost

Engineering Contradiction:

Improveinformation accessibilityVSAvoidsearch time

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The system performs preliminary object identification and information retrieval automatically before the user needs it. AI models continuously analyze video streams to detect objects, and the system pre-loads relevant information so that when a user shows interest in an object, the information is immediately available without requiring manual search actions.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system serves itself by automatically identifying objects in the video stream and retrieving their associated information without user intervention. The AI-driven system monitors video content, identifies objects of interest, and proactively provides information services, eliminating the need for users to manually search for object-related information.

Inventive Principle:

Principle #25Self-service

2Loss of information

If manual searching is performed to find information about objects in video streams, then information can be obtained, but user focus is broken

Engineering Contradiction:

Improveinformation accessibilityVSAvoiduser focus continuity

Core Design Contradiction:

Loss of informationVSEase of operation

Solution Approach 1:

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The AI-driven information service acts as an intermediary between the user and the vast amount of information available about objects in the video stream. Instead of users directly searching through multiple sources, the intermediary system automatically identifies objects and retrieves relevant information, simplifying the interaction and maintaining user focus on the video content.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Measurement precision

If prior knowledge is required to search for object information, then specific information can be found, but knowledge barriers are created

Engineering Contradiction:

Improvesearch accuracyVSAvoiduser accessibility

Core Design Contradiction:

Measurement precisionVSAdaptability or versatility

Solution Approach 1:

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system changes the parameter of information retrieval from user-initiated to system-initiated. By transforming the search process into an automated AI-driven task, the system adapts to users regardless of their knowledge level, providing accurate object identification and information retrieval without requiring users to possess specific search skills or domain knowledge.

Inventive Principle:

Principle #35Parameter changes

4Productivity

If AI models are used to identify objects in real-time, then user-specific information services are provided, but system complexity increases

Engineering Contradiction:

Improveinformation service efficiencyVSAvoidsystem complexity

Core Design Contradiction:

ProductivityVSDevice complexity

Solution Approach 1:

The system segments the complex AI-driven information service into distinct modular components: video stream processing module, object detection module using AI models, information retrieval module, and user-specific service delivery module. Each component handles a specific task independently, making the overall complex system manageable and maintainable while achieving high productivity in providing user-specific information services.

Inventive Principle:

Principle #1Segmentation

Data Source

PatentUS12489953B2Systems and methods for leveraging machine learning to enable user-specific real-time information services for identifiable objects within a video stream

Publication Date: 2025.12.02 ADEIA GUIDES INC

AI summary

A first content type of a first media stream being output is determined, and at least one first classification model is selected based on the first content type. Objects are identified in the first media stream using the at least one first classification model and information related to the objects is generated for presentation. A change in output of media streams from the first media stream to a second media stream may then be detected. A second content type of the second media stream is then determined and at least one second classification model is selected based on the second content type. Objects are then identified in the second media stream using the at least one second classification model and information related to the objects in generated for presentation.