Media Event Generation Using Episode Segmentation and ML Scoring

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Users face difficulty in organizing large collections of media, such as photos and videos, due to the lack of effective methods for identifying and categorizing events within their media libraries.

Innovation Solution

A machine-learning model is employed to segment media libraries into episodes and identify events based on likelihood and significance, using a trained event machine-learning model that generates an event signal and significance score, allowing for the display of media in a user interface organized by events.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If manual review and curation of search results is performed, then result accuracy and relevance are improved, but time consumption and operational costs increase

Engineering Contradiction:
Improveresult accuracyVSAvoidtime consumption
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent replaces manual review and curation of search results with an automated machine learning model. The model automatically generates events from search results, eliminating the need for human operators to manually review and curate results, thus reducing time consumption while maintaining result accuracy through automated processing.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The system enables self-service by allowing the event generation process to occur automatically without human intervention. The machine learning model autonomously processes search results, generates events, and updates the event store, making the system self-sufficient and eliminating dependency on manual operational services.

Inventive Principle:
Principle #25Self-service

2Reliability

If comprehensive search queries are executed to ensure all relevant events are captured, then event detection completeness is improved, but computational resources and processing time increase

Engineering Contradiction:
Improveevent detection completenessVSAvoidcomputational resources
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The system performs preliminary actions by proactively monitoring and detecting events before they become critical. The machine learning model continuously analyzes search results and generates events in advance, allowing the system to prepare and respond to potential issues before they impact operations, thereby improving detection completeness without requiring exhaustive real-time searching.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent employs parameter changes by dynamically adjusting search and processing parameters based on event criticality and system state. The machine learning model optimizes search query parameters, result filtering thresholds, and processing intensity to balance detection completeness with computational resource consumption, ensuring efficient resource utilization while maintaining reliable event detection.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP4204992B1Automatic generation of events using a machine-learning model
Publication Date: 2026.05.06 GOOGLE LLC
  • EP4204992B1 patent drawingFigure 1
  • EP4204992B1 patent drawingFigure 2
  • EP4204992B1 patent drawingFigure 3

AI summary

A media application segments a library of media associated with a user account into episodes, wherein each episode is associated with a corresponding time period. The media application generates, using an event machine-learning model, an event signal that indicates a likelihood that an event occurred in each episode, wherein the event machine-learning model is a classifier that receives the media as input. The media application generates an event significance score for each episode. The media application determines one or more events from the episodes based on the event signal and a corresponding event significance score exceeding a threshold event significance value. The media application provides a user interface that includes corresponding media from the one or more events.