TouchID: User Authentication on Mobile Devices via Inertial-Touch Gesture Analysis
summary
Combines on-screen touch data and phone inertial sensors to authenticate users by analyzing how a graphic unlock pattern is performed, implemented on commercial smartphones and evaluated with volunteers. Uses spatial alignment and gesture segmentation to reduce intra-user variance, a four-part feature selection (start/end/turns/paths) with Fisher Score for uniform representation, and a multi-threshold kNN to tolerate posture changes. Achieves a 4.9% equal error rate and resists mimicking attacks.
AI-written summary · about
related talks
06mListen to Your Fingers: User Authentication based on Geometry Biometrics of Touch Gesture
19mEvaluating the Influence of Targets and Hand Postures on Touch-based Behavioural Biometrics
20mBiometric Touch Sensing: Seamlessly Augmenting Each Touch with Continuous Authentication
45mYour passport to the 7th continent
19mThumprint: Socially-Inclusive Local Group Authentication Through Shared Secret Knocks
19mStay Cool! Understanding Thermal Attacks on Mobile-based User Authentication
06mTowards Usable and Secure Location-based Smartphone Authentication
22mThe high cost of cheap biometrics