Eye Tracking Latency Reduction Through Cascaded Eye Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing eye-tracking technologies face challenges in accurately extracting biometric information due to eyelid occlusions and variations in eye appearance, leading to inefficiencies in gaze estimation and biometric identification.
Innovation Solution
A detailed eye shape model is estimated using cascaded shape regression techniques, which enhance the detection of eye features by identifying the boundaries of the pupil, iris, and eyelids, allowing for more robust gaze estimation and biometric identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional eye-tracking methods are used to extract biometric information, then the process can be performed with simpler algorithms, but accuracy is reduced due to eyelid occlusions and variations in eye appearance
Solution Approach 1:
The patent applies preliminary action by performing cascaded shape regression to estimate eye shape (pupil, iris, eyelid boundaries) before extracting biometric features. This preliminary modeling of eye anatomy prepares the data structure needed for accurate feature extraction, allowing the system to handle eyelid occlusions and appearance variations systematically before the actual biometric measurement occurs.
2Measurement precision
If the entire eye image is searched for feature detection, then all possible features can be identified, but processing time and computational resources increase significantly
Solution Approach 1:
The patent applies segmentation by dividing the eye image into distinct anatomical regions (pupil, iris, eyelids) using shape regression models. This segmentation allows the system to focus feature detection and gaze estimation computations only within the relevant segmented regions rather than searching the entire eye image, significantly reducing processing time while maintaining detection completeness.
Solution Approach 2:
The patent applies local quality by applying different processing strategies to different parts of the eye based on the shape regression model. The system treats the pupil, iris, and eyelid regions with region-specific algorithms appropriate to their anatomical characteristics, improving both detection accuracy and efficiency by avoiding uniform processing of the entire image.
3Productivity
If simple eye models are used, then processing is faster and simpler, but the models cannot handle variations in eye appearance and eyelid occlusions effectively
Solution Approach 1:
The patent applies dynamics by using a cascaded shape regression approach that adaptively models eye shape based on actual image content rather than assuming fixed geometric patterns. The model dynamically adjusts to different eye appearances, eyelid positions, and occlusion scenarios, allowing the system to maintain processing efficiency while handling diverse anatomical variations.
Data Source
AI summary
Systems and methods for eye tracking latency enhancements. An example head-mounted system obtains a first image of an eye of a user. The first image is provided as input to a machine learning model which has been trained to generate iris and pupil segmentation data given an image of an eye. A second image of the eye is obtained. A set of locations in the second image at which one or more glints are shown is detected based on iris segmentation data generated for the first image. A region of the second image at which the pupil of the eye of the user is shown is identified based on pupil segmentation data generated for the first image. A pose of the eye of the user is determined based on the detected set of glint locations in the second image and the identified region of the second image.