kaushi019 / facial-recognition-using-facenet-siamese-one-shot-learning

Implementation of Facial Recognition System Using Facenet based on One Shot Learning Using Siamese Networks

Facial-Recognition-Using-FaceNet-Siamese-One-Shot-Learning

This program has been used to implement Facial Recognition using Siamese Network architecture. The implementation of the project is based on the research paper :

FaceNet: A Unified Embedding for Face Recognition and Clustering arXiv:1503.03832 by Florian Schroff, Dmitry Kalenichenko, James Philbin

Facenet implements concept of Triplet Loss function to minimize the distance between anchor and positive images, and increase the distance between anchor and negative images.

Prerequisites

h5py==2.8.0
Keras==2.2.4
tensorflow==1.13.0rc2
dlib==19.16.0
opencv_python==3.4.3.18
imutils==0.5.1
numpy==1.15.2
matplotlib==3.0.0
scipy==1.1.0

Install the packages using pip install -r requirements.txt

Usage

To use the facial recognition system, you need to have a database of images through which the model will calculate image embeddings and show the output vector. The images which are in the database are stored as .jpg files in the directory ./images.

To generate your own dataset and add more faces to the system, use the following procedure:

Sit in front of your webcam. Use the Image_Dataset_Generator.py script to save 50 images of your face. Use this command: python Image_Dataset_Generator.py to generate images which will be saved in images folder.

To use the facial recognition system, run the command on your terminal : python face_recognizer.py

References

1. The code has been implemented using deeplearning.ai course Convolutional Networks Week 4 Assignment, which has the files fr_utils.py and inception_blocks_v2.py
2. The keras implementation of the model is by Victor Sy Wang's implementation and was loaded using his code: https://github.com/iwantooxxoox/Keras-OpenFace.