BlinkListener: “Listen” to Your Eye Blink Using Your Smartphone
summary
System uses smartphone speaker and microphone to emit FMCW acoustic chirps, separate eye reflection via range bins, and model I-Q vectors so eyelid blinks appear as a bump on a head-movement arc distinct from phase-only breathing motion. Implementations on Bela and iPhone 5c run a real-time detector that fits the arc to choose an optimal viewing position and detect sparse blinks, achieving a median detection accuracy of 95% across handheld, glasses, and crowded scenarios.
AI-written summary · about