Audio-Guided Video Selection for Social Monitoring

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

The challenge lies in efficiently selecting meaningful videos from a large number of recordings made by users during events, as excessive video transmission can overwhelm communication networks and incur unnecessary data costs, especially when videos contain insignificant information.

Innovation Solution

A method and system that analyze audio content captured by audio capture components in conjunction with video content to determine if a defined event exists, and then assess whether the image capture component captured event-related video content based on the direction of the event relative to the audio capture component, allowing for the selection and presentation of relevant video and audio content.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If all user videos are transmitted to the centralized center for monitoring, then complete event coverage is achieved, but network bandwidth is overloaded and data costs increase

Engineering Contradiction:

Improveevent coverage completenessVSAvoidnetwork bandwidth consumption

Core Design Contradiction:

ReliabilityVSQuantity of substance

Solution Approach 1:

The system performs preliminary audio analysis on user videos before transmission to the centralized center. The audio content is examined to detect defined events, and only videos containing such events are selected for transmission. This preliminary filtering action prevents network overload by reducing the volume of transmitted data while ensuring that all significant events are captured and monitored.

Inventive Principle:

Principle #10Preliminary action

2Reliability

If all user videos are transmitted to the centralized center for monitoring, then complete event coverage is achieved, but data usage costs increase

Engineering Contradiction:

Improveevent coverage completenessVSAvoiddata usage cost

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

Inventive Principle:

Principle #10Preliminary action

3Loss of information

If manual video selection is performed from many recordings, then meaningful videos can be identified, but time and operational complexity increase

Engineering Contradiction:

Improvevideo content qualityVSAvoidvideo review time

Core Design Contradiction:

Loss of informationVSLoss of time

Solution Approach 1:

The system replaces manual video review with automated audio analysis. The audio content of each video is automatically examined to detect defined events, and videos containing such events are automatically selected. This mechanical substitution eliminates the time-consuming manual review process while maintaining high-quality video selection through objective audio-based event detection.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

4Measurement precision

If direction analysis is performed to verify event-related video content, then video relevance accuracy is improved, but computational complexity increases

Engineering Contradiction:

Improvevideo event relevance accuracyVSAvoidanalysis system complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The system uses audio content as an intermediary to verify video event relevance. Instead of directly analyzing video content for event detection, the system first analyzes the audio track to detect defined events, then uses the audio event information to guide video content verification through direction analysis. This intermediary approach improves accuracy while managing computational complexity by leveraging the simpler audio analysis as a filtering step.

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS10726270B2Selecting media from mass social monitoring devices

Publication Date: 2020.07.28 FEC IP LLC

AI summary

Technologies are generally described that relate to managing videos relating to events. An example technique can include analyzing audio content captured by an audio capture component and associated with video content captured by an image capture component to facilitate determining whether a defined event exists in the audio content. The technique also can include, in response to determining that the defined event exists in the audio content based on the analyzing, determining whether the image capture component captured event-related video content related to the defined event based on a result of analyzing a first direction in which the defined event occurred relative to the audio capture component in relation to a second direction that the image capture component was facing while capturing the video content, in accordance with a defined match criterion, to facilitate determining whether to present, emphasize, or select at least the video content.