A lightweight computer-vision based web authentication project that demonstrates passwordless login using face recognition and real-time user presence checks.
FaceAuth v2 is a browser-based facial login application built to showcase how computer vision can be used for:
- user face enrollment
- real-time recognition
- live user action confirmation
- protected session access
The project removes dependency on traditional passwords and introduces an interactive visual identity verification flow.
✅ Webcam face registration
✅ Real-time facial login verification
✅ Blink + mouth movement live checks
✅ Timed session expiration
✅ Request throttling for repeated attempts
✅ Lightweight Flask architecture
✅ No heavy deep learning framework dependency
| Layer | Technology |
|---|---|
| Backend | Flask |
| Recognition Engine | OpenCV LBPH |
| Detection | Haar Cascades |
| Numerical Processing | NumPy |
| Frontend | HTML / CSS / JavaScript |
pip3 install -r requirements.txt
python3 app.pyRun locally at:
http://127.0.0.1:5000User captures a facial sample through webcam and the model stores the trained identity.
User performs:
- blink detection
- mouth-open verification
- face recognition match
If matched successfully, access is granted to the dashboard.
facial_auth_v2/
├── app.py
├── face_engine.py
├── requirements.txt
├── data/
├── templates/
└── static/- Smart attendance systems
- Lab identity verification demos
- Passwordless local portals
- Educational computer vision projects
- Modern authentication experiments
Unlike many facial login demos, this implementation is:
- lightweight
- easy to run locally
- Python 3.14 compatible
- framework independent
- suitable for practical academic and portfolio demonstrations
- multi-user database integration
- admin panel
- deployment containerization
- audit logs
- improved UI analytics
Computer Vision meets Modern Authentication.