Hybrid Eye Tracker Using DVS Camera for Low-Power AR
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current eye tracking technologies in AR glasses require high processing power, leading to battery strain and high costs, and struggle with object recognition in low-light conditions due to the need for continuous illumination and high frame rates.
Innovation Solution
The use of Dynamic Vision Sensor (DVS) cameras in conjunction with CMOS cameras for eye tracking, allowing for reliable, low-power, and wide dynamic range eye movement detection without additional lighting, utilizing DVS cameras' ability to capture changes in pixel intensity for gaze tracking and blink detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional cameras with illumination systems are used for eye tracking, then object recognition accuracy is improved, but power consumption increases and battery life decreases
Solution Approach 1:
The patent replaces conventional frame-based camera systems with an event-based Dynamic Vision Sensor (DVS) camera that operates on a different detection principle. Instead of capturing continuous frames requiring illumination, the DVS camera detects changes in pixel intensity asynchronously, enabling eye tracking without continuous illumination and reducing power consumption while maintaining tracking accuracy
Solution Approach 2:
The system uses periodic illumination only when needed (e.g., during initialization or when intensity changes are detected) rather than continuous illumination. The DVS camera's event-driven architecture naturally implements periodic data capture based on actual changes in the scene, reducing overall power consumption while maintaining measurement precision
2Reliability
If conventional cameras with illumination systems are used for eye tracking, then dynamic range is improved, but device complexity and cost increase
Solution Approach 1:
The patent substitutes the complex illumination system and frame-based processing architecture with a simpler DVS camera system that inherently provides wide dynamic range through its event-based detection mechanism. The DVS camera's ability to detect intensity changes across a wide range without additional lighting simplifies the overall system architecture while maintaining or improving dynamic range performance
Solution Approach 2:
The patent extracts and removes the illumination system from the eye tracking setup by using the DVS camera's ability to function without it. This extraction simplifies the device complexity and reduces cost while the DVS camera's inherent wide dynamic range capabilities maintain reliable eye tracking performance
3Speed
If conventional cameras are used for eye tracking, then frame rate can be increased for accurate tracking, but processing power requirements increase and battery life decreases
Solution Approach 1:
The patent replaces the conventional frame-based high-speed camera system with an event-based DVS camera that achieves equivalent or superior tracking performance with significantly lower processing requirements. The DVS camera's asynchronous event stream requires less computational power to process while maintaining the speed necessary for accurate eye tracking
Solution Approach 2:
Instead of capturing complete frames at high rates, the DVS camera captures only the necessary intensity change events at the moment they occur. This partial action approach (capturing only changes rather than full frames) reduces processing power requirements while maintaining the effective speed needed for accurate eye tracking
4Device complexity
If conventional cameras without illumination are used to reduce cost, then device complexity is reduced, but object recognition in dark environments becomes difficult
Solution Approach 1:
The patent uses the DVS camera's unique intensity change detection mechanism to replace systems that would require illumination for dark environment operation. The DVS camera can detect eye movements and features in dark environments by sensing intensity changes without needing additional light sources, maintaining measurement precision while keeping the system simple and cost-effective
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
This solution enables efficient eye tracking with reduced power consumption, extended battery life, and effective object recognition in darker environments, providing a cost-effective and immersive AR experience.
Implementation Method 1
Dynamic Vision Sensor (DVS) cameras... utilize DVS cameras' ability to capture changes in pixel intensity
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system and method enable an electronic device to perform gaze tracking operations. The electronic device includes one or more cameras and at least one processor. The one or more cameras include a Dynamic Vision Sensor (DVS) camera. The one or more cameras are configured to capture features of an eye of a wearer of the electronic device The processor is configured to receive, from the one or more cameras, an image or pixel stream associated with the eye, determine an initial pose of the eye based on the image or pixel stream, receive DVS data from the DVS camera, track one or more changes to the initial pose of the eye based on the DVS data, and based on the one or more changes to the initial pose of the eye, determine a gaze of the wearer.