SmileAuth: Using Dental Edge Biometrics for User Authentication on Smartphones
summary
SmileAuth implements a camera-based Android prototype that captures a few images while the phone moves and extracts dental edge lines via Sobel edge detection after mouth segmentation and noise removal, then computes four feature types and learns feature weights with a random forest while estimating camera angle from mouth landmark distances. Evaluated on over 300 volunteers, it reports 99.74% precision, 98.69% F-score, 0.25% FPR and 2.31% FNR and resists image, video, denture and forced-attacks.
AI-written summary · about
chapters (8)
- 0:00Intro
- 0:35Background
- 0:53Motivation
- 1:39Proposed Method
- 2:43System Overview
- 3:07Smile Auth Design
- 4:13Basic Experiment Setup
- 5:10Conclusion
related talks
45mYour passport to the 7th continent
06mListen to Your Fingers: User Authentication based on Geometry Biometrics of Touch Gesture
40mNew Face, Who Dis? Protecting Privacy in a World of Surveillance
45mOverview of Biometric Auth
06mTowards Usable and Secure Location-based Smartphone Authentication
22mThe high cost of cheap biometrics
26mVirtual U: Defeating Face Liveness Detection by Building Virtual Models...
14mTransforming Digital Identity with AI