This repository is a part of the project: DOneLogin: Single Sign On (SSO) Provider combined with Facial Two-Factors Authentication (2FA).
It contains 3 repositories:
- oidc_provider_with_facial_mfa: SSO Provider Web service
- facial_verification_android: An Android application to support 2FA with facial verification feature.
- MobileFaceNet_TF: MobileFaceNet deep face recognition model
For more detail about this project, visit this project description.
This repository is the implementation of Android application which support facial 2FA.
It applies deep learning techniques for face detection/recognition/anti-spoofing on frames captured from smart phone's frontal camera.
- Face detection: Face detector of Google's MLKit on Android.
- Face recognition: MobileFaceNet (source code | paper)
- Face anti-spoofing: Deep Tree Learning for Zero-shot Face Anti-Spoofing (pre-train | source code | paper)
The pre-train of face anti-spoofing model has many problems, but it works in many cases. I'll find another model/method later.
- Linking to accounts on SSO web service
- Face detection
- Face recognition
- Face anti-spoofing
- Notifications powered by FCM
- View/Accept/Reject face authentication request
- Remove accounts
- Sync authentication requests with SSO web service
- Sync account removal with SSO web service
- Improve MobileFaceNet with new SOTA model
- Improve face anti-spoofing model with CDCD (source code | paper)
This project is inspired by the following repositories: