Eye Tracking With Background 3D Eye Model Updating
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing eye tracking and modeling technologies require extensive user interaction and enrollment processes, which are inconvenient for users wanting immediate extended reality experiences.
Innovation Solution
Simultaneous tracking of eye characteristics and updating of an eye model using infrared light reflections and image sensors, allowing for minimal user interaction and background enrollment during normal activities.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If separate techniques are used for eye tracking and eye model generation, then measurement precision can be maintained, but device complexity and user interaction requirements increase
Solution Approach 1:
The patent combines eye tracking and eye model generation into a single integrated system that processes image frames simultaneously for both purposes. The same image capture hardware and processing pipeline serve dual functions: tracking eye characteristics (pupil position, gaze direction) and building/updating the 3D eye model, eliminating the need for separate enrollment and tracking processes.
Solution Approach 2:
The system makes the image processing system universal by using it for multiple functions: initial eye model capture, continuous eye model updates, and real-time eye characteristic tracking. The same computational routines analyze image frames for both model generation and tracking, allowing the system to perform multiple tasks with a single unified apparatus.
2Measurement precision
If extensive enrollment processes with multiple look instructions are implemented, then eye model accuracy improves, but user convenience and ease of operation deteriorate
Solution Approach 1:
The system enables self-service enrollment where the user's eye naturally moves through the field of view during normal activities, and the system automatically captures sufficient data to build the eye model without requiring the user to follow specific instructions or perform deliberate eye movements. The coverage map tracks which portions of the eye have been captured and automatically determines when enrollment is complete.
Solution Approach 2:
The system performs preliminary capture of eye images during natural user activity before formal tracking begins. By continuously capturing and evaluating image frames against the coverage map during normal use, the system prepares the eye model in advance, so that when tracking is needed, the model is already available and accurate.
3Productivity
If continuous eye tracking and model updating are performed simultaneously, then productivity and speed improve, but computational resource requirements and processing complexity increase
Solution Approach 1:
The system applies partial action by selectively processing only those image frames that contribute new information to the eye model, as determined by comparing against the coverage map. Rather than processing every frame for both tracking and model building, the system identifies and processes only the subset of frames that advance model completion, reducing redundant computation while maintaining accuracy.
Solution Approach 2:
The system dynamically adjusts its processing based on the current state of model completion. As the coverage map shows which portions of the eye have been captured, the system adapts its frame selection and processing priorities, intensifying model-building processing when coverage is incomplete and focusing more on tracking when the model is established, optimizing computational resources over time.
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
Enables seamless eye tracking and modeling without dedicated enrollment processes, providing efficient and unobtrusive eye characteristic determination and three-dimensional model generation.
Implementation Method 1
directing light towards an eye of a user using one or more light (e.g., IR) sources to produce glint reflections
Implementation Method 2
obtains, via one or more sensors (e.g., at least one image sensor), image frames comprising depictions of reflections of the light via the eye
Data Source
AI summary
Various implementations disclosed herein include an inconspicuous eye enrollment for user devices through the simultaneous or near-simultaneous eye characteristic tracking and eye model updating. The implementations disclosed herein direct light towards an eye of a user to produce glint reflections and obtain, via one or more sensors, image frames comprising depictions of the glint reflections. Based on those image frames, the implementations disclosed herein track eye characteristics and update an eye model of the eye during a period of time between obtaining a first image frame of the image frames and obtaining a second image frame of the image frames.