Eye Enrollment Triggering with Guest Gaze Gestures
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional gaze-based interactions in extended reality systems fail to effectively initiate eye enrollment for guest users, as their eye models are not known, making it difficult for them to use gaze-based user interfaces.
Innovation Solution
Implementing gesture-based methods, such as rolling the eyes in a circle or moving them randomly, to trigger partial or full eye enrollment, allowing the system to collect gaze data in the background and perform visual axis or full eye enrollment based on data coverage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If conventional gaze-based UI gestures are used for eye enrollment, then gaze-based interactions can be maintained, but guest users cannot initiate enrollment because their eye models are not known and gaze interactions do not work well
Solution Approach 1:
The system performs preliminary background collection of gaze data and corneal reflection images automatically before the guest user needs to enroll. This preliminary data collection enables the system to recognize eye gestures and initiate enrollment without requiring the user to perform complex gaze-based UI interactions first.
Solution Approach 2:
The system uses an intermediary gesture recognition mechanism that translates physical eye movements (gestures) into enrollment commands. This intermediary layer allows guest users to initiate enrollment through simple eye gestures rather than requiring precise gaze-based UI interactions that would normally be needed.
2Reliability
If full eye enrollment is performed for guest users, then gaze-based interactions work well, but the enrollment process becomes complex and time-consuming
Solution Approach 1:
The system performs partial eye enrollment (visual axis enrollment) as a minimum requirement for guest users, which is sufficient for basic gaze-based interactions. Full eye model enrollment is performed only when necessary and when background data collection has provided adequate coverage, avoiding unnecessary complexity for most guest user scenarios.
Solution Approach 2:
The system dynamically adjusts the enrollment process based on the quality and coverage of background collected data. When background data is sufficient, the system requires fewer enrollment steps. When background data is insufficient, the system automatically extends the enrollment process to collect additional parameters, thereby adapting the complexity level to the actual needs.
3Ease of operation
If gaze data is collected in the background, then eye enrollment can be initiated with simple gestures, but additional data collection time and processing requirements are introduced
Solution Approach 1:
The system continuously collects gaze data and corneal reflection images in the background during normal device operation, ensuring that data is always available when a guest user needs to enroll. This continuous background collection eliminates the need for separate data collection phases and integrates enrollment initiation seamlessly into the user experience.
Solution Approach 2:
The system performs preliminary background data collection automatically before the guest user initiates enrollment. This preliminary action ensures that sufficient data is already available to support gesture recognition and enrollment initiation, reducing the overall time required for the enrollment process.
Data Source
AI summary
Methods to trigger eye enrollment without requiring good gaze interaction are described that allow a guest user of a device to initiate partial or full eye enrollment even though their eye model is not known and thus conventional gaze-based interactions do not work well. A gaze tracking system collects gaze data in the background. At any time (or within an interval after a user puts on the device), an eye enrollment can be triggered by detecting some gaze gesture, for example rolling the eyes in a large circle, or moving the eyes randomly for a time that exceeds a threshold. Depending on the coverage of the gaze/cornea data collected in the background, a full eye enrollment or only a visual axis enrollment may be performed in response to the gesture.