Gaze and Head Pose Validation for Accurate XR Object Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing XR environments face challenges in accurately determining user interactions with virtual objects due to false positives from relying solely on gaze inputs, leading to degraded user experience and increased power consumption.
Innovation Solution
Combining gaze vectors with head pose information to confirm user interactions, using head motions to validate selections and reduce the need for additional inputs, thereby improving accuracy and reducing power consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If gaze input alone is used to determine user interaction, then device complexity is reduced, but interaction accuracy deteriorates due to false positives
Solution Approach 1:
The patent combines multiple input modalities (gaze input and head pose input) into a unified interaction determination system. The processor determines user interaction by evaluating both gaze vectors and head pose values together, merging these separate sensing channels to achieve more accurate interaction detection while maintaining manageable system complexity through integrated processing.
Solution Approach 2:
The patent introduces head pose information as an intermediary validation mechanism between gaze input and final interaction determination. The system uses head pose values as a mediating factor to confirm or reject gaze-based interaction attempts, thereby improving accuracy without requiring a complete redesign of the interaction architecture.
2Measurement precision
If additional input validation is added to improve accuracy, then interaction accuracy is improved, but power consumption increases
Solution Approach 1:
The patent implements periodic sampling of head pose data at specific intervals rather than continuous monitoring. The system samples head pose values periodically to validate gaze-based interactions, which reduces the cumulative power consumption compared to continuous monitoring while maintaining sufficient accuracy for interaction determination.
Solution Approach 2:
The system uses partial validation by checking head pose values only when needed to confirm or reject gaze-based interactions, rather than continuously processing all sensor data. This selective processing approach reduces power consumption while maintaining the necessary accuracy for reliable interaction detection.
3Measurement precision
If head pose validation is implemented, then interaction accuracy is improved, but device complexity increases
Solution Approach 1:
The patent makes the existing gaze-tracking and head pose sensing systems serve multiple functions: gaze vectors are used for both initial interaction detection and head pose data is used for both orientation understanding and interaction validation. This multi-functionality approach improves interaction accuracy without proportionally increasing device complexity.
Solution Approach 2:
The system uses the head pose sensing infrastructure already present in the device to provide validation functionality. The existing sensors and processing pathways are leveraged to create the validation mechanism, avoiding the need for entirely separate dedicated validation hardware and reducing the complexity increase that would otherwise be required.
Data Source
AI summary
Various implementations disclosed herein include devices, systems, and methods for using a gaze vector and head pose information to effectuate a user interaction with a virtual object. In some implementations, a device includes a sensor for sensing a head pose of a user, a display, one or more processors, and a memory. In various implementations, a method includes displaying a set of virtual objects. Based on a gaze vector, it is determined that a gaze of the user is directed to a first virtual object of the set of virtual objects. A head pose value corresponding to the head pose of the user is obtained. An action relative to the first virtual object is performed based on the head pose value satisfying a head pose criterion.