Estimation of Gaze Detection Accuracy Using the Calibration Information-Based Fuzzy System

Area of Science:

• Human-Computer Interaction
• Computer Vision
• Biomedical Engineering

Background:

• Gaze tracking systems rely on initial calibration to map eye coordinates to screen positions.
• Calibration accounts for individual eye variations and the kappa angle.
• Previous research focused on calibration methods but not on estimating accuracy from calibration data.

Purpose of the Study:

• To propose a novel method for estimating the final accuracy of a gaze tracking system.
• To utilize user calibration information within a fuzzy system for accuracy prediction.
• To validate the proposed method across different calibration markers and display techniques.

Main Methods:

• Developed a fuzzy system to estimate gaze tracking accuracy.
• Integrated user calibration data into the fuzzy system.
• Conducted experiments using a near-infrared camera with four marker types and three display methods.

Main Results:

• The proposed fuzzy system accurately estimated gaze tracking accuracy.
• Accuracy estimation was consistent across various marker types and display methods.
• The method provides a reliable measure of gaze detection performance post-calibration.

Conclusions:

• A fuzzy system effectively estimates gaze tracking accuracy using calibration data.
• The method offers a robust solution for predicting system performance.
• This approach enhances the reliability of gaze tracking applications.