juantrader500/linux-lockscreen-bioauth-faceid-gestures- — explained in plain English
Analysis updated 2026-05-18
Unlock a Fedora GNOME lock screen with your face and a thumbs up gesture instead of typing a password.
Add a PAM-based biometric authentication step that safely falls back to password login on failure.
Run offline, CPU-only face recognition using InsightFace without needing a GPU.
| juantrader500/linux-lockscreen-bioauth-faceid-gestures- | 0-bingwu-0/live-interpreter | 0cm-labs/tokenizer-benchmark | |
|---|---|---|---|
| Stars | 2 | 2 | 2 |
| Language | Python | Python | Python |
| Setup difficulty | hard | moderate | moderate |
| Complexity | 5/5 | 2/5 | 2/5 |
| Audience | ops devops | general | researcher |
Figures from each repo's GitHub metadata at analysis time.
Requires editing PAM configuration files directly, a mistake can lock you out of your own login.
This project adds two-factor biometric login to Fedora Linux systems running GNOME. Instead of only typing a password at the lock screen, the system can recognize your face and check for a specific hand gesture, and only if that combination fails or times out does it fall back to asking for your ordinary password. It works by plugging into PAM, the standard Linux system that handles logins. When GDM, the GNOME login manager, asks for credentials, PAM runs this tool first. If your face is recognized and you make a thumbs up gesture, you get in without typing anything, otherwise the normal password prompt appears after a short timeout. Face recognition is handled by a model called InsightFace, running on the CPU rather than requiring a graphics card. The first time it runs, it looks at a folder of your own reference photos, front, left, and right, and turns each one into a numeric representation of your face, which it saves to a cache so future logins are almost instant. Live camera frames are then compared against that cache using a similarity score, and a close enough match counts as a face match. The cache automatically refreshes itself if you add, remove, or change any of the reference photos. The gesture check uses a separate tool called MediaPipe to track the points of your hand in real time, and looks for a thumbs up: the thumb clearly raised above the wrist while the other four fingers stay folded down. Both the face and the gesture checks have to succeed for a few frames in a row before access is granted, which helps avoid a false match from one odd camera frame. The project is written in Python and is tested on Fedora 44 with GNOME, though the author says it should also work on similar distributions like RHEL, CentOS, and Rocky Linux, with partial support for others such as Ubuntu, Debian, Arch, and Linux Mint. Setting it up involves creating an isolated Python environment, installing a specific set of libraries, providing 10 to 16 photos of your own face, and then editing PAM configuration files directly, which the README warns should always be backed up first since a mistake there can lock you out of your own system.
A two-factor biometric login system for Fedora GNOME that unlocks the screen using face recognition and a hand gesture, falling back to your normal password if it fails.
Mainly Python. The stack also includes Python, InsightFace, MediaPipe.
Setup difficulty is rated hard, with roughly 1h+ to a first successful run.
Mainly ops devops.
This repo across BitVibe Labs
Don't trust strangers blindly. Verify against the repo.