ElectraSight: Fully Onboard Eye Tracking for Smart Glasses With Hybrid EOG (hEOG)

Smart glasses with integrated eye tracking technology are revolutionizing diverse fields, from immersive augmented reality experiences to cutting-edge health monitoring solutions. However, traditional eye tracking systems rely heavily on cameras and significant computational power, leading to high-energy demand and privacy issues. Alternatively, systems based on electrooculography (EOG) provide superior battery life but are less accurate and primarily effective for detecting blinks, while being highly invasive. To bridge this gap, the paper introduces ElectraSight, a system built upon a new concept we define as hybrid Electrooculography (hEOG). This approach combines contact and contactless electrodes to create a robust, low-power, and truly non-invasive eye tracking system. To validate our approach, we collected a comprehensive dataset from 20 participants, using a commercial eye-tracker for ground-truth labeling. A lightweight 1D Convolutional Neural Network (CNN), quantized to 4-bit and occupying just 79 kB of memory, performs real-time eye movement classification. Without requiring user-specific calibration, the model achieves 81% accuracy for 10 classes and 92% for 6 classes. Experimental results demonstrate that ElectraSight delivers high accuracy in eye movement and blink classification, with minimal overall movement detection latency (90% within 60ms) and an ultra-low inference time (301 ms). The power consumption settles down to 7.75mW for continuous data acquisition and 46μJ for the tinyML inference. This efficiency enables continuous operation for over 3 days on a compact 175 m A h battery. This work opens new possibilities for eye tracking in commercial applications, offering an unobtrusive solution that enables advancements in user interfaces, health diagnostics, and hands-free control systems.