Thumprint: Socially-Inclusive Local Group Authentication Through Shared Secret Knocks
summary
Shared one-secret knock that still identifies individuals using an instrumented surface with accelerometer and microphone signals, DSP features (Fourier and Daubechies wavelets), correlation feature selection, per-user clustering and template matching; implemented as a prototype that gates a laptop or smart lock. Two-day lab study with 30 participants shows 24-hour consistency and that casual adversaries are mostly detected, yielding about a 12% equal error rate with under 5% misidentification — a promising proof of concept.
AI-written summary · about
chapters (7)
- 0:00Start
- 4:07Thumprint Overview
- 4:47Registration
- 5:09Supervised Feature Selection
- 6:11Training
- 6:56Authentication
- 11:06Evaluation
related talks
06mListen to Your Fingers: User Authentication based on Geometry Biometrics of Touch Gesture
06mTouchID: User Authentication on Mobile Devices via Inertial-Touch Gesture Analysis
06mVocalLock: Sensing Vocal Tract for Passphrase-Independent User Authentication Leveraging ...
22mComplementing Passkeys with Strong Identity Assurance for High-Risk Actions and Account Recovery
24mSecurity at your fingertips: Theory → Practice!
21mMaking authentication your competitive advantage under GDPR and PSD2
25mGreat Privacy Challenge for Reusable Identity
06mTowards Usable and Secure Location-based Smartphone Authentication