Skip to content

A simple general purpose embedding generator for facial recognition.

Notifications You must be signed in to change notification settings

GriffinCanCode/facekeeper

Repository files navigation

Facekeeper - Face Encoding Generator

A simple tool to capture and encode human faces into 128-dimensional numpy arrays for use in facial recognition applications.

Overview

This program provides a user-friendly way to:

  • Capture face images through a camera
  • Process and validate face quality
  • Generate standardized 128-D face encodings
  • Store encodings with associated names
  • Preview and verify captured faces
  • Export encodings for use in other applications
  • Scare your friends who don't know how to use this program or what it does

The face encodings are generated using dlib's deep learning model and can be used for:

  • Face verification/matching
  • Identity management systems
  • Access control applications
  • Custom facial recognition projects

Features

  • Live camera preview with face detection
  • Quality checks for optimal face capture
  • Multiple sample collection for accuracy
  • Both GUI and CLI interfaces
  • Configurable capture parameters
  • Secure local storage of encodings
  • Export options for encoded data

Setup

  1. Install required dependencies:
pip install -r requirements.txt
  1. Download dlib models:
python setup.download_models.py
  1. Run the program:
python main.py

The program focuses on generating high-quality, consistent face encodings that can be reliably used for facial recognition tasks in other applications. Totally open source and free, let me know if you have any questions or suggestions.

About

A simple general purpose embedding generator for facial recognition.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages