BioAuthenticator is a secure "Intent-Based" vault for iOS that reimagines biometric security. Unlike standard FaceID which is passive, this app requires a specific, user-defined facial expression (the "Secret Look") to decrypt sensitive data.
Standard biometrics can be bypassed if a phone is held up to a sleeping or unwilling user's face. BioAuthenticator solves this by requiring active intent. The user must perform a specific gestureβlike a wink, a smile, or a cheek puffβto "unlock" the encrypted vault.
- ARKit Face Tracking: Monitors 52 unique facial blend shapes (muscle movements) at 60fps.
- Biometric Mapping: Maps facial geometry to a
[String: Float]dictionary for precise gesture matching. - Haptic Engine (Taptic): Uses
UIImpactFeedbackGeneratorto provide tactile confirmation of a successful biometric handshake. - SwiftUI + Liquid Glass UI: Features a high-fidelity "blurred vault" aesthetic that "melts" away only upon successful verification.
- Privacy-First Architecture: No biometric data is sent to a server; all facial snapshots are stored locally on-device.
- Enrollment: The user performs their "Secret Look" while the TrueDepth camera maps the coordinates of their facial muscles.
- Persistence: The gesture is saved to the local secure storage as a mathematical threshold.
- Verification: When entering the vault, the app compares the real-time
ARFaceAnchordata against the saved key. - Decryption: If the "Match Score" exceeds 80%, the vault is unblurred and haptic feedback is triggered.
- Frameworks: SwiftUI, ARKit, SceneKit, Combine
- Hardware: iPhone TrueDepth Camera System (iPhone X or newer)
- Language: Swift 6.0
- Deployment: iOS 18.0+
Sakshar Dhawan