jonnyoo/glance

What it solves

Glance provides a FaceID-like unlock experience for MacBooks, allowing users to unlock their computers using facial recognition instead of typing a password or using TouchID.

How it works

The app uses Apple's Vision and Core ML frameworks to process camera frames on-device. During enrollment, it captures the user's face from multiple angles to create a 512-dimensional mathematical embedding (a fingerprint) using an ArcFace-based model. When the Mac wakes or locks, the app scans for a matching face and performs liveness checks to ensure it is a real person rather than a photo. Once verified, it uses macOS Accessibility permissions to automatically type the stored, encrypted password into the lock screen.

Who it’s for

Mac users (macOS 15+) who want a more convenient, hands-free way to unlock their devices and are comfortable with the security trade-offs of 2D camera-based recognition.

Highlights

• On-device processing: All face data and passwords are encrypted and stored locally; nothing is sent to the internet.
• Liveness detection: Includes "Light" and "Heavy" modes to detect spoofs like printed photos or phone screens by analyzing motion and reflections.
• Notch integration: Features a dynamic island-style UI built into the MacBook notch with fluid animations.
• Multiple identities: Supports enrolling several people or different versions of the same person (e.g., with glasses or a beard).
• Privacy-focused: Original camera images are discarded immediately after embeddings are created.