GhostID: Enabling Non-Persistent User Differentiation in Frequency-Division Capacitive ...
summary
GhostID extracts capacitive ghosting artifacts from a frequency-division multi-touch controller to tell whether simultaneous touches come from the same hand, the same person’s two hands, or different people. A research prototype uses an opaque capacitive sensor and frequency-division transmit/receive rows, runs at 60 fps, and was tested with 21 participants over 600,000+ frames. It achieves about 98% precision for distinguishing users and drops to ~77% reliability for same-user two-hand classification; it needs at least two touches and shows grounding limitations.
AI-written summary · about