Eye-Tracking Gaze Estimation With Confidence-Based Single-Eye Fallback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current eye-tracking technologies face challenges in accurately determining gaze information due to the compact layout of wearable devices, which results in suboptimal camera angles and low accuracy in capturing eye gaze directions.
Innovation Solution
The method involves determining target feature points in acquired eye images, assessing confidence levels, and using a gaze estimation model to combine gaze information from images with varying confidence levels, optimizing the shooting angle of the image capture apparatus, and enabling gaze information determination based on single-eye data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the camera is positioned in a compact wearable device layout, then the device can be worn comfortably, but the camera angle becomes suboptimal resulting in low accuracy for capturing eye gaze directions
Solution Approach 1:
The patent transitions from traditional 2D eye image analysis to 3D gaze direction estimation by introducing depth information through a gaze estimation model. The model processes eye images to output three-dimensional gaze direction data (azimuth and elevation angles), effectively adding a dimensional aspect to the measurement that compensates for suboptimal camera angles in compact wearable devices.
2Measurement precision
If traditional eye-tracking methods are used with compact device layout, then device portability is maintained, but gaze information determination accuracy becomes low
Solution Approach 1:
The patent changes the fundamental parameters of gaze information representation from traditional 2D coordinates to 3D spherical coordinates (azimuth and elevation angles). This parameter transformation enables more accurate gaze determination by accounting for the vertical component of gaze direction, which is particularly important when cameras are positioned at suboptimal angles in compact wearable devices.
3Volume of moving object
If eye images are captured at large angles in compact devices, then device size is reduced, but the captured eye gaze direction becomes far from the camera resulting in low accuracy
Solution Approach 1:
The patent introduces a gaze estimation model as an intermediary computational layer between the camera capture and the final gaze direction output. This model acts as a mediator that corrects for the geometric distortions and angular deviations inherent in compact device layouts, transforming suboptimal camera angle data into accurate three-dimensional gaze direction estimates.
Data Source
AI summary
A method and apparatus for determining gaze information, and an eye-tracking device are provided. The method includes: acquiring two eye images; determining their respective target feature points and their corresponding respective confidence levels; determining first gaze information of a second eye image based on the target feature points of the second eye image when the confidence level corresponding to a first eye image of two eye images is less than a preset confidence level and that the confidence level corresponding to the second eye image is greater than or equal to the preset confidence level, determining the first gaze information as first gaze information of the first eye image; inputting the two eye images respectively into a gaze estimation model to obtain respective second gaze information; and determining target gaze information corresponding to the two eye images based on the first gaze information and the second gaze information.