Facial Recognition Authentication (Siamese Network)

This project implements a Facial Recognition Authentication System using a Siamese Network.
It uniquely identifies individuals during authentication activities like login by learning similarity functions.

• Model: Siamese Network (TensorFlow/Keras)
• Dataset: LFW - Labeled Faces in the Wild
• Reference Paper: Koch et al. - Siamese Neural Networks for One-Shot Learning

Installation

pip install -r requirements.txt

Run Locally

python application.py
# open http://localhost:5000

Deploy on Heroku/Render

• Ensure Procfile and requirements.txt are included.
• Add your trained siamese_model.h5 at the project root.
• Deploy 🚀

Usage

• Upload two face images.
• The system outputs a similarity score between 0 and 1.
• Higher scores mean higher similarity.