faceverifyai 0.1.0

Siamese CNN based face verification package

FaceVerifyAI

A Siamese CNN-based face verification Python package built with TensorFlow/Keras for checking whether two face images belong to the same person.

This package provides a simple API for face similarity verification using a pretrained deep learning model with a custom L2Normalize layer.

✨ Features

• Verify whether two face images match
• Siamese CNN architecture for similarity learning
• Easy Python API (verify_face)
• Installable as a reusable Python package
• Tested on local machine + Google Colab
• Includes custom TensorFlow layer support
• Automatic image preprocessing

⚙️ Installation

Install from GitHub

git clone https://github.com/abhisekkundu-DS/siamese-face-verification.git
cd siamese-face-verification
pip install -e .

pip install git+https://github.com/abhisekkundu-DS/siamese-face-verification.git

Create a virtual environment

python -m venv tfenv
Windows PowerShell:

powershell
.\tfenv\Scripts\Activate
Linux/Mac:

bash
source tfenv/bin/activate
 Install the package
bash
pip install -e .

Example :

Create a test file
Create test.py in the root folder:

python
from faceverifyai import verify_face

result = verify_face("person1.jpg", "person2.jpg")
print(result)

Run the test
bash
python test.py
 Expected output
python
{'match': True, 'score': 0.94}