Screen-Content Gaze Calibration Without Explicit User Setup
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing gaze tracking systems often require time-consuming and resource-intensive personalized calibration, which is not optimal for various devices and user constraints, leading to inaccuracies in determining what a user is looking at on a display screen.
Innovation Solution
An implicit calibration method using spatiotemporal information and a neural network to generate a personalized gaze function based on displayed content and uncalibrated gaze information, without explicit calibration steps, allowing for real-time calibration on individual user devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If explicit personalized calibration is performed using research-grade eye tracker, then measurement precision of gaze tracking is improved, but loss of time and resource consumption increase significantly
Solution Approach 1:
The system performs preliminary calibration actions by collecting gaze data from multiple users during normal device operation and pre-computing calibration parameters. This allows the calibration model to be ready in advance, eliminating the need for time-consuming calibration sessions for each individual user while maintaining accurate gaze tracking.
Solution Approach 2:
The patent creates a generalized calibration model by copying and adapting calibration data from multiple users. Instead of performing explicit calibration for each individual, the system uses aggregated gaze patterns from a population to generate a calibration model that can be applied to new users, significantly reducing calibration time while preserving measurement precision.
2Measurement precision
If explicit personalized calibration is performed using research-grade eye tracker, then measurement precision of gaze tracking is improved, but device complexity and resource requirements increase
Solution Approach 1:
The patent extracts the calibration function from complex research-grade eye tracking systems and implements it using simpler, more accessible technology. By separating the calibration data collection and processing into a standalone model that can run on standard devices, the system achieves research-grade accuracy without requiring complex hardware or extensive calibration procedures.
Solution Approach 2:
The system changes the parameters of the calibration approach by using aggregated gaze data from multiple users and applying machine learning techniques to generate calibration models. This transforms the calibration process from a hardware-intensive, user-specific procedure into a software-based, population-derived solution that reduces device complexity while maintaining precision.
3Measurement precision
If traditional calibration approaches are used, then personalized gaze tracking is achieved, but productivity and ease of operation decrease due to multiple training scenarios
Solution Approach 1:
The patent creates a universal calibration model that serves multiple users and multiple device types simultaneously. The aggregated gaze data and resulting calibration model can be applied across different users and devices without requiring separate calibration procedures for each case, significantly improving productivity and ease of operation while maintaining personalized tracking accuracy.
Solution Approach 2:
The system enables self-service calibration by automatically collecting and processing gaze data during normal device usage. Users do not need to participate in explicit calibration sessions or follow training scenarios; the calibration model is generated and applied automatically in the background, improving both productivity and user experience.
Data Source
AI summary
The technology relates to methods and systems for implicit calibration for gaze tracking. This can include receiving, by a neural network module, display content that is associated with presentation on a display screen (1202). The neural network module may also receive uncalibrated gaze information, in which the uncalibrated gaze information includes an uncalibrated gaze trajectory that is associated with a viewer gaze of the display content on the display screen (1204). A selected function is applied by the neural network module to the uncalibrated gaze information and the display content to generate a user-specific gaze function (1206). The user-specific gaze function has one or more personalized parameters. And the neural network module can then apply the user-specific gaze function to the uncalibrated gaze information to generate calibrated gaze information associated with the display content on the display screen (1208). Training and testing information may alternatively be created for implicit gaze calibration (1000).