GAPses: Versatile Smart Glasses for Comfortable and Fully-Dry Acquisition and Parallel Ultra-Low-Power Processing of EEG and EOG
Abstract
GAPses is an unobtrusive smart-glasses platform that captures EEG and EOG using fully dry soft electrodes and processes data on-board using a parallel ultra-low-power GAP9 RISC-V processor. The design eliminates continuous wireless streaming, enhancing privacy and reliability, and supports multiple interaction tasks including alpha-wave measurement, steady-state visual evoked potentials, motor-movement classification, and biometric recognition. EEG-based subject identification reaches 98.87% sensitivity and 99.86% specificity using eight EEG channels with 121 μJ per inference, while an EOG eye-movement classifier achieves 96.68% accuracy at 24 μJ per inference. Total system power of 16.28 mW enables more than 12 hours of continuous operation on a 75 mAh battery.
Key Highlights
- Fully dry soft electrodes integrated into glasses capture both EEG and EOG with on-board GAP9 processing.
- Achieves 98.87% sensitivity / 99.86% specificity for EEG-based identification and 96.68% accuracy for EOG eye-movement classification.
- Consumes as little as 121 μJ per inference and 16.28 mW total system power, enabling 12+ hours of continuous operation.
Thorir Mar Ingolfsson
Postdoctoral Researcher
I develop efficient machine learning systems for biomedical wearables that operate under extreme resource constraints. My work bridges foundation models, neural architecture design, and edge deployment to enable real-time biosignal analysis on microwatt-scale devices.