Multi-Object AR Content Generation via Intent Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for generating augmented reality (AR) content in electronic devices are limited to single-human interactions and do not effectively consider multi-object interactions and intents in real-world environments, resulting in inaccurate insertion of AR text and effects.
Innovation Solution
An electronic device method that identifies and classifies the posture, action, intent, and interaction of multiple objects in a scene using a Machine Learning model, generating AR content by aggregating these interactions over time and mapping them to predefined templates with real-time information from a knowledge base.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing methods use single-human gesture recognition for AR content insertion, then the system complexity remains low, but the accuracy of AR content insertion deteriorates in multi-object environments
Solution Approach 1:
The system segments the scene into multiple objects and individually identifies postures, actions, intents, and interactions for each object. This segmentation allows the system to handle multi-object environments by processing each object separately, thereby improving AR content insertion accuracy without overwhelming system complexity
Solution Approach 2:
The patent introduces an intermediary processing layer that aggregates interactions and intents of multiple objects over time before generating AR content. This intermediary step transforms complex multi-object data into structured information that can be mapped to AR templates, resolving the contradiction between handling multiple objects and maintaining system simplicity
2Measurement precision
If the system aggregates interactions and intents of multiple objects over time, then the accuracy of AR content generation improves, but the processing time increases
Solution Approach 1:
The system performs preliminary classification of postures and actions for each object, and pre-aggregates their interactions and intents over time. This preliminary processing organizes the data structure in advance, enabling faster AR content generation when needed while maintaining high accuracy through comprehensive temporal aggregation
Solution Approach 2:
The patent implements periodic aggregation of object interactions and intents at defined time intervals rather than continuous processing. This periodic approach maintains accurate temporal understanding of multi-object behaviors while reducing computational burden and processing time compared to continuous analysis
Data Source
AI summary
An electronic device and a method for generating an augmented reality (AR) content in an electronic device are provided. The method includes determining a posture and an action of each object of the plurality of the objects in the scene displayed on a field of view of the electronic device, classifying the posture and the action of each object of the plurality of the objects in the scene, identifying an intent and an interaction of each object from the plurality of objects in the scene based on at least one of the classified posture and the classified action, and generating the AR content for the at least one object in the scene of at least one of the identified intent and the identified interaction of the at least one object.