MemX: An Attention-Aware Smart Eyewear System for Personalized Moment Auto-capture
summary
MemX is a retrofit smart-eyewear system that auto-captures personal moments by combining a lightweight continuous eye tracker with an on-demand scene analyzer running a temporal visual attention deep model. It uses an inward-facing eye camera and a forward-facing world camera to fuse gaze and scene features, triggering expensive analysis only on detected attention events. Evaluation uses YouTube-VIS and 30 participants plus 11 field pilots, reports 93% average precision for moment detection, and achieves 8 hours of continuous operation on a single charge.
AI-written summary · about
related talks
20mEye-tracking and AI: Unveiling the Limitations and Synergies
15mMeCap: Whole-Body Digitization for Low-Cost VR/AR Headsets
20mBuilding a Personalized, Auto-Calibrating Eye Tracker from User Interactions
17mMadEye: Boosting Live Video Analytics Accuracy with Adaptive Camera Configurations
19mEveryday Eye Contact Detection Using Unsupervised Gaze Target Discovery
19mEyeScout: Active Eye Tracking for Position and Movement Independent Gaze Interaction with Large ...
15mC-Face: Continuously reconstructing facial expressions by deep learning contours of the face ...
33mSee More, Be More