XR Intent Prediction Using Visual Hot-Region Motion Cues
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing user interaction methods in extended reality (XR) environments, such as VR, AR, and MR, suffer from low accuracy and stability, requiring multiple selection operations, leading to user fatigue and dissatisfaction due to inefficiency.
Innovation Solution
A method and apparatus that utilize a visual indicator identifier to predict user interactive intention by determining the relative position and motion features of the identifier within predicted hot regions of interaction objects, allowing for quicker and more accurate selection of target objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traditional gesture control mode or eye movement control mode is used, then user can interact with virtual world, but user needs to perform multiple selection operations leading to time waste and energy consumption
Solution Approach 1:
The system performs preliminary action by predicting the user's intended target before the user completes the selection operation. The prediction model analyzes eye movement trajectory and gesture motion features to forecast the target object in advance, allowing the system to prepare for the selection operation beforehand. This reduces the time and energy required for multiple selection operations by anticipating the user's intent.
2Measurement precision
If traditional control modes are used, then user can select target objects, but selection accuracy is unstable requiring multiple operations
Solution Approach 1:
The system implements feedback by continuously monitoring eye movement data and gesture motion features, comparing the predicted target with the actual user selection, and using this information to refine future predictions. The prediction model receives feedback from the user's actual selection behavior and adjusts its parameters accordingly, improving both selection accuracy and stability over time.
3Ease of operation
If multiple selection operations are performed, then user can select desired target, but user fatigue increases leading to poor interaction experience
Solution Approach 1:
The system performs self-service by automatically predicting and highlighting the target object based on eye movement and gesture analysis, reducing the burden on the user to manually perform multiple selection operations. The system serves itself by anticipating user intent and preparing the interface accordingly, allowing users to complete selections with fewer, less exhausting operations.
Data Source
AI summary
A method and an apparatus for predicting user interactive intention, and a storage medium are provided. The method includes: acquiring position information of a visual indicator identifier, the visual indicator identifier being configured to mark a manipulation position of an user in a preset space; determining a relative position relationship between the visual indicator identifier and a predicted hot region of an interaction object according to the position information of the visual indicator identifier; determining a first motion feature of the visual indicator identifier in the predicted hot region, when it is determined that the visual indicator identifier is located in the predicted hot region according to the relative position relationship; and determining the interaction object as a target interaction object with which the user has interactive intention when the first motion feature of the visual indicator identifier matches with an intra-hot-region motion feature.