3D Eye Model Gaze Correction Using Historical Inaccuracy Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing techniques for estimating and correcting eye-based attributes with respect to viewing content on a display of a device lack simplicity and accuracy, leading to undesirable viewing experiences.
Innovation Solution
Implementing a 3D eye model-based prediction system that accounts for historical inaccuracies using a correction process, incorporating sensor data from enrollment and real-time user sessions to adjust gaze predictions with a transformer model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing techniques are used for estimating and correcting eye-based attributes, then the system implementation is straightforward, but the accuracy of gaze position estimation is insufficient
Solution Approach 1:
The gaze estimation system is segmented into two distinct components: an enrolled eye model for initial prediction and a correction process for accuracy improvement. This segmentation allows each component to specialize in specific tasks, with the enrolled eye model handling broad gaze prediction and the correction process refining accuracy by compensating for historical inaccuracies, thereby achieving high measurement precision without requiring complete system redesign
Solution Approach 2:
A correction process acts as an intermediary between the enrolled eye model predictions and the final gaze position output. This intermediary component receives predictions from the enrolled eye model, applies corrections based on historical inaccuracy data, and produces the final corrected gaze position, thereby improving accuracy without directly modifying the enrolled eye model structure
2Measurement precision
If a 3D eye model is used for gaze prediction, then the theoretical accuracy is high, but the model inaccuracy accumulates over time without correction
Solution Approach 1:
The system implements feedback by using historical inaccuracy information from the enrolled eye model predictions to correct future predictions. The correction process continuously learns from past errors and applies compensatory adjustments, creating a closed-loop system that maintains prediction reliability over time by constantly refining accuracy based on accumulated performance data
Solution Approach 2:
The enrolled eye model performs preliminary gaze position prediction before the correction process refines it. This preliminary action provides an initial estimate that guides subsequent correction steps, allowing the system to establish a baseline prediction that can then be optimized with historical inaccuracy data to achieve final high accuracy
Data Source
AI summary
Various implementations disclosed herein include devices, systems, and methods that predict a gaze position. For example, a process may obtain an enrolled eye model of a user that was determined based on sensor data obtained via one or more sensors of an electronic device. The process may further obtain eye-model inaccuracy information corresponding to inaccuracies of prior gaze position predictions determined using the enrolled eye 3D model and generate a predicted gaze position based on the enrolled eye model. The predicted gaze position may correspond to a position on display of the electronic device. The process may further generate a corrected gaze position based on output of a correction process that receives as input the predicted gaze position and the eye-model inaccuracy information.