Eye Movement Characterization Using Scene-Based Reference Frames
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing eye tracking technologies struggle to accurately characterize eye movements in dynamic scenarios, such as those involving vestibulo-ocular reflex, smooth pursuit, and optokinetic nystagmus, due to the challenges posed by head movements and changing environments.
Innovation Solution
A method and system that utilize a head-wearable device with scene and eye cameras to generate sequences of images, determining motion signals for the scene and eye, calculating a difference signal to characterize eye movements, and classifying them based on this difference, allowing for reliable characterization even in scenarios with head movement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If threshold-based algorithms (dispersion or velocity methods) are used to detect fixations, then the detection works well in head-fixed scenarios, but it becomes unreliable in freely moving scenarios due to VOR, smooth pursuit, and optokinetic nystagmus
Solution Approach 1:
The patent introduces an intermediary reference frame (scene-based coordinate system) that mediates between the eye movement signals and the fixation detection algorithm. By transforming gaze points into this intermediate reference frame that moves with the head and scene, the system enables reliable fixation detection in freely moving scenarios while maintaining compatibility with traditional threshold-based methods
Solution Approach 2:
The patent changes the reference frame parameter from a fixed head-mounted coordinate system to a dynamic scene-based coordinate system. This parameter change allows the same threshold-based algorithms to work reliably across different head movement scenarios by adapting the reference frame to account for VOR, smooth pursuit, and optokinetic nystagmus
2Adaptability or versatility
If head-wearable eye tracking devices are used, then the range of application scenarios is expanded to include outdoor activities and tasks requiring free head movement, but the accuracy of eye movement characterization deteriorates in dynamic scenarios
Solution Approach 1:
The patent introduces a scene-based reference frame as an intermediary that connects the head-wearable device measurements with the actual fixation events. This intermediary reference frame, which accounts for head motion and scene motion, enables accurate eye movement characterization across diverse application scenarios including outdoor activities and tasks requiring free head movement
3Device complexity
If traditional fixation detection algorithms are used in freely moving scenarios, then the computational simplicity is maintained, but the detection accuracy decreases due to large dispersion and velocity of gaze points during VOR, smooth pursuit, and optokinetic nystagmus
Solution Approach 1:
The patent introduces a scene-based reference frame transformation as an intermediary step that preserves the simplicity of threshold-based algorithms while improving accuracy. The transformation itself is computationally straightforward (coordinate system change), but it enables accurate fixation detection by accounting for head and scene motion, thus maintaining algorithmic simplicity while achieving high measurement precision
Data Source
Figure 1A~1B
Figure 2A~2B
Figure 3A~3B
AI summary
A method for characterizing eye movements of a user includes generating a sequence of scene images of a scene in a field of view of the user and a corresponding sequence of eye images of at least a portion of an eye of the user, determining a first signal referring to a motion, and using the sequence of eye images to determine a second signal referring to a motion of the eye of the user, determining a difference signal corresponding to a difference between the second signal and the first signal, and determining a characteristic of the eye movements based on the difference signal.