Eye Tracker Calibration via Automatic Region of Attraction Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional eye tracker calibration methods require user cooperation and attention, leading to inefficiencies and stress, especially in applications involving multiple users or dynamic environments, and do not provide continuous or optimized calibration.
Innovation Solution
A calibration method and device that automatically determines optimal calibration functions by analyzing user eye movements without user interaction, using a selection module to identify regions of attraction, a calculation module to generate calibration positions, and an evaluation module to select the best calibration function, ensuring continuous adaptation to user and environmental changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional calibration methods are used requiring user cooperation and attention, then calibration can be performed, but user stress and time consumption increase
Solution Approach 1:
The system performs calibration automatically without user intervention. The calibration device autonomously tracks eye movements and computes calibration functions while the user simply views normal content, eliminating the need for user cooperation during calibration while maintaining accuracy
Solution Approach 2:
The system continuously performs calibration in the background before actual use. By maintaining ongoing calibration processes, the system ensures calibration is already optimized when needed, preventing user stress during critical moments
2Reliability
If periodic calibration is performed to compensate for environmental changes and eye fatigue, then calibration accuracy is maintained, but user time and collaboration requirements increase
Solution Approach 1:
The system performs calibration continuously in the background rather than periodically interrupting the user. The calibration process runs continuously while the user views content, maintaining up-to-date calibration data without requiring dedicated calibration sessions
Solution Approach 2:
The system maintains calibration in advance through continuous background processing. By constantly updating calibration functions, the system ensures optimal calibration is always ready without requiring last-minute user intervention
3Adaptability or versatility
If manual calibration requiring user cooperation is used, then calibration can be performed, but it becomes unacceptable for applications with multiple users or dynamic environments
Solution Approach 1:
The calibration system operates autonomously without requiring user cooperation. This self-service approach allows seamless transitions between multiple users and adapts to dynamic environments without interrupting any user, making it suitable for shared and public applications
Solution Approach 2:
The system designed to work universally across multiple users and applications. By eliminating user-specific calibration requirements, the same system serves diverse users in various contexts including public displays, shared devices, and dynamic environments
4Ease of operation
If calibration is made imperceptible to the user, then user effort is minimized, but calibration precision and reliability must be maintained
Solution Approach 1:
The system performs calibration autonomously in the background without user awareness. The calibration device independently tracks eye movements and computes calibration functions, making the process imperceptible while maintaining precision through continuous automated measurement
Solution Approach 2:
The system continuously monitors eye position and uses feedback to refine calibration functions in real-time. This ongoing feedback loop ensures high precision is maintained even though the user is unaware of the calibration process
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A calibration method for an eye tracker (2) provides for the defining of at least one first region of attraction (A1; A2 ) of a first image to be displayed to a user; acquiring a first sequence of data relative to the eye movement of a user who is looking at the first image by means of an eye tracker (2); calculating at least a plurality of first calibration positions by means of respective calibration functions (fj) on the basis of a first gaze determined from the first sequence of data relative to eye movement; assigning a first score to each calibration function (fj) on the basis of the respective first calibration position and the first region of attraction (A1 A2 ); and selecting one of the calibration functions (fj) on the basis of the score.