Attention-Based XR Content Visualization Reducing Cognitive Load
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Extended reality systems face challenges in providing user interfaces that effectively present content to users, particularly in environments with a large quantity of information, leading to clutter and reduced usability due to overwhelming cognitive loads.
Innovation Solution
A computer-implemented method that adapts content visualization in extended reality environments by inferring user interest through egocentric vision input data, using rule-based or machine learning AI to prioritize and modify virtual content relevance, thereby promoting more important content and reducing clutter.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If all virtual content is displayed in the extended reality environment, then the user receives complete information, but the interface becomes cluttered and usability decreases
Solution Approach 1:
The patent applies local quality by making different parts of the interface have different visual prominence based on their relevance to the user's current attention. Virtual content elements are rendered with varying sizes, opacities, or visual weights according to their relevance score, which is determined by the machine learning model's prediction of user interaction likelihood. This allows important information to stand out while less important information remains visible but subdued, resolving the contradiction between information completeness and usability.
Solution Approach 2:
The patent implements dynamics by continuously adapting the visual presentation of virtual content based on real-time analysis of user attention and interaction patterns. The machine learning model dynamically adjusts the relevance and prominence of different content elements as the user moves through the environment, ensuring that the interface remains optimized for current user needs rather than presenting a static cluttered display.
2Loss of information
If multiple virtual content elements are displayed simultaneously, then comprehensive information is provided, but cognitive load increases and user understanding decreases
Solution Approach 1:
The system applies local quality by varying the visual prominence of different content elements based on their predicted relevance to user interaction. The machine learning model assigns different weights to various virtual content elements, and the rendering system translates these weights into visual differences such as size, brightness, or positional emphasis. This allows the interface to maintain information completeness while reducing cognitive load by making the hierarchy of importance visually apparent.
3Adaptability or versatility
If the interface presents all available functions and information, then full functionality is accessible, but the user interface becomes cumbersome and difficult to use
Solution Approach 1:
The patent implements dynamics by continuously adapting the interface presentation based on real-time analysis of user attention, gaze patterns, and interaction history. The machine learning model predicts which functions and information elements are most relevant at any given moment and adjusts their visual prominence accordingly. This dynamic adaptation allows the interface to maintain full functionality while presenting only the most relevant elements with high prominence, reducing cognitive load and improving ease of operation.
Solution Approach 2:
The system applies self-service by using the user's own attention patterns and interaction behavior to automatically adjust the interface presentation. The machine learning model learns from user behavior and autonomously determines which content elements should be emphasized, eliminating the need for manual interface configuration and ensuring the interface adapts to individual user preferences and needs.
Data Source
AI summary
Techniques for adaptively visualizing content in an artificial environment based on the attention of a user. In one particular aspect, a computer-implemented method is provided that includes obtaining input data from a user, inferring content that is of interest to the user based on features gathered from the user's attention in the input data, identifying virtual content data based on the content that is of interest to the user, determining modifications to be applied to the virtual content data based on relevancy, applying the modifications to the virtual content data to generate a final format for the virtual content data, and rendering virtual content in the extended reality environment displayed to the user based on the final format for the virtual content data. The virtual content rendered from relevant virtual content data is more prominently displayed as compared to the virtual content rendered from semi-relevant and non-relevant virtual content data.