Media Trend Detection Using Pose Embeddings

Solution Overview

Problem

Content sharing platforms face challenges in efficiently detecting and maintaining media trends due to the resource-intensive and inaccurate nature of identifying media items associated with trends based on user-provided metadata, leading to wasted computing resources and increased latency.

Innovation Solution

A system that utilizes audiovisual embeddings and pose embeddings to identify media items with common characteristics, calculating distance scores and coherence scores to determine media trends, leveraging AI models for accurate and efficient trend detection and maintenance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the platform detects media trends and identifies associated media items using traditional methods, then media trends can be identified, but the process is time consuming and resource intensive

Engineering Contradiction:

Improvetrend detection accuracyVSAvoiddetection time

Core Design Contradiction:

Measurement precisionVSLoss of time

Solution Approach 1:

The system pre-computes and stores pose embeddings for media items in advance. When trend detection is needed, these pre-computed embeddings are readily available for rapid comparison and analysis, eliminating the need for real-time pose estimation and significantly reducing detection time while maintaining accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces traditional metadata-based trend detection with AI-powered pose embedding comparison. This substitution enables automatic, efficient analysis of media item characteristics without manual intervention, reducing both time and resource requirements while improving detection precision

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If the platform detects media trends and identifies associated media items using traditional methods, then media trends can be identified, but it is resource intensive

Engineering Contradiction:

Improvetrend detection accuracyVSAvoidcomputing resource consumption

Core Design Contradiction:

Measurement precisionVSUse of energy by moving object

Solution Approach 1:

The system pre-computes and stores pose embeddings for media items in advance. When trend detection is needed, these pre-computed embeddings are readily available for rapid comparison and analysis, eliminating the need for real-time pose estimation and significantly reducing detection time while maintaining accuracy

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces traditional metadata-based trend detection with AI-powered pose embedding comparison. This substitution enables automatic, efficient analysis of media item characteristics without manual intervention, reducing both time and resource requirements while improving detection precision

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Productivity

If the platform uses user-provided metadata to identify media items associated with trends, then trend detection can be performed, but accuracy is reduced

Engineering Contradiction:

Improvedetection efficiencyVSAvoidtrend identification accuracy

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The patent replaces traditional metadata-based trend detection with AI-powered pose embedding comparison. This substitution enables automatic, efficient analysis of media item characteristics without manual intervention, reducing both time and resource requirements while improving detection precision

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system transitions from analyzing user-provided metadata to analyzing pose embeddings derived from media item content. This parameter change enables more accurate and objective trend identification by focusing on actual visual characteristics rather than potentially inaccurate or incomplete user metadata

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS20250111675A1Media trend detection and maintenance at a content sharing platform

Publication Date: 2025.04.03 GOOGLE LLC

AI summary

Methods and systems for media trend detection and maintenance are provided herein. A set of media items each having common media characteristics is identified. A set of pose values is determined for each respective media item of the set of media items. Each pose value is associated with a particular predefined pose for objects depicted by the set of media items. A set of distance scores is calculated. Each distance score represents a distance between the respective set of pose values determined for a media item and a respective set of pose values determined for an additional media item. A coherence score is determined for the set of media items based on the calculated set of distance scores. Responsive to a determination that the coherence score satisfies one or more coherence criteria, a determination is made that the set of media items corresponds to a media trend of a platform.