This project combines the best ideas from Skuldur/facenet-face-recognition and Sefik Ilkin Serengil. Thank you for writing awesome code and tutorials!
This repository contains a demonstration of face recognition using the FaceNet network and one-shot learning. It uses a pretrained model to encode the images and uses Euclidean distance to find the difference. If you need a quick introduction to FaceNet and one-shot learning, refer this: Making your own Face Recognition System.
To install all the requirements for the project run
pip install -r requirements.txt
Download the model, facenet_keras.h5 and put it in models
directory.
Keep the face images of people you want to recognize in databases
directory.
Keep the input photo (on which you want to run the program) in input
directory.
In the root directory. After the modules have been installed you can run the project by using python
python facenet.py
And see the outputs in output
directory