HMD Eye Tracking Calibration Using Head and Eye Rotation Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing head-mounted displays (HMDs) face challenges in accurately calibrating eye tracking systems, which are crucial for providing effective virtual, augmented, and mixed reality experiences, due to variations in head and eye movements that affect gaze direction estimation.
Innovation Solution
An HMD system incorporates an eye-tracking assembly with an inertial measurement unit and a camera to track head and eye rotations, using machine-learning models to refine calibration by comparing eye and head rotations, and predict future gaze directions based on image analysis and light source illumination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional eye tracking calibration methods are used in head-mounted displays, then the system structure remains simple, but gaze direction estimation accuracy deteriorates due to variations in head and eye movements
Solution Approach 1:
The patent combines the eye tracking assembly with the head tracking system by integrating an inertial measurement unit (IMU) to measure head rotation and comparing it with eye rotation data from the eye tracking assembly. This merging of head and eye tracking functions allows the system to compensate for head movement variations and improve gaze direction estimation accuracy without requiring completely separate calibration systems.
Solution Approach 2:
The system implements feedback by continuously comparing eye rotation measurements with head rotation measurements from the IMU. This feedback mechanism allows the calibration state to be dynamically refined based on the relationship between eye and head movements, improving measurement accuracy while maintaining a unified system structure.
2Measurement precision
If dynamic calibration refinement is implemented by comparing eye and head rotation, then gaze estimation accuracy improves, but processing complexity and computational requirements increase
Solution Approach 1:
The system performs preliminary calibration by establishing the relationship between eye rotation and head rotation during initial setup and during dynamic operation. By pre-establishing this calibration model, the system reduces the computational complexity of real-time gaze estimation, as the comparison between eye and head rotation data follows a predetermined calibration framework rather than requiring complex real-time calculations.
3Measurement precision
If multiple sensors including IMU and camera are integrated for simultaneous head and eye tracking, then measurement accuracy improves, but device weight and power consumption increase
Solution Approach 1:
The eye tracking assembly serves multiple functions: it tracks eye rotation for gaze direction estimation and simultaneously provides feedback for calibrating the relationship between eye and head movements. The integrated system uses the same camera and light source for both eye tracking and calibration purposes, reducing the need for separate dedicated components and thereby minimizing additional weight.
4Measurement precision
If continuous calibration refinement is performed during use, then long-term measurement accuracy is maintained, but energy consumption increases
Solution Approach 1:
The system performs calibration refinement periodically and dynamically during operation by comparing eye rotation with head rotation data from the IMU. This periodic calibration approach maintains measurement accuracy over time without requiring continuous intensive processing, thereby managing energy consumption more efficiently compared to constant high-power calibration modes.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The system enhances the accuracy of gaze direction estimation, allowing for improved interaction and rendering of virtual environments by aligning display content with user gaze and head movements.
Implementation Method 1
tracking head rotation of a user during a time period while the head-mounted display is worn by the user; the head rotation is measured using an inertial measurement unit in the head-mounted display
Implementation Method 2
acquiring an image of the eye, using a camera mounted in the head-mounted display, while the eye is illuminated using the light source
Implementation Method 3
the camera images light from the light source in the infrared
Implementation Method 4
illuminating an eye, using a light source mounted in the head-mounted display; the camera images light from the light source in the infrared
Data Source
AI summary
Calibration for eye tracking calibration in a head-mounted display can refined during use by tracking head rotation of a user during a time period while the head-mounted display is worn by the user, tracking eye rotation of the user during the time period, comparing the eye rotation of the user to the head rotation of the user during the time period, and refining a calibration state of eye tracking for the head-mounted display based on comparing the eye rotation of the user to the head rotation of the user during the time period.