phucvinh57/biopass

An alternative to Windows Hello/Howdy

What it solves

Linux lacks a unified, modern biometric login equivalent to Windows Hello. Existing solutions like Howdy offer only face recognition and rely on command-line configuration. Biopass fills this gap by providing both face and fingerprint authentication with a modern GUI for management.

How it works

Biopass uses EdgeFace for face recognition and YOLO-Face for face detection. It supports anti-spoofing through an IR camera and an embedded AI model. Fingerprint authentication is also included. Authentication methods can be registered and run in parallel or sequentially. The tool provides a GUI for configuration and management, and integrates with PAM for system sign-in and Polkit for interactive authentication.

Who it's for

Linux users who want a more complete biometric login experience, especially those who prefer a GUI over command-line tools and need both face and fingerprint support.

Highlights

  • Supports both face and fingerprint authentication
  • Modern GUI for management and configuration
  • Anti-spoofing via IR camera and embedded AI model
  • Can run authentication methods in parallel or sequentially
  • Integrates with PAM and Polkit
  • Available via Debian, RPM, and AUR packages

Related

  • Project
    egdels/makeacopy

    MakeACopy is an offline-first Android document scanner that uses local machine learning models for edge detection and OCR to create searchable PDFs without any cloud dependency.

  • Project
    dweep-desai/FaceGate-Mac

    A native macOS app-locker that uses on-device face recognition, Touch ID, or passwords to restrict access to specific applications.

  • Project
    zhenyingfang/Awesome-Temporal-Action-Detection-Temporal-Action-Proposal-Generation

    An “awesome\-list” style resource\. It collects research papers on detecting human actions in videos \(finding when actions start/end\) plus several side\-tasks: generating candidate time\-windows \(proposals\), training with weaker labels\, online detection\, and open\-vocabulary detection\. Papers are listed by publication year with links to papers and sometimes code\.

  • Project
    Labellerr/Hands-On-Learning-in-Computer-Vision

    Labellerr Notebooks is a collection of practical Jupyter notebooks that show how to fine-tune computer vision models like YOLO for real-world tasks such as waste classification, traffic analysis, sports analytics, and industrial inspection. Each notebook is paired with a YouTube tutorial and can be run in Google Colab.