Siamese Network for Image Pair Discrimination

Project Overview

This project involves the development of a Siamese neural network designed to assess the similarity between pairs of images. Leveraging TensorFlow and Keras, the network utilizes convolutional neural layers to extract features from each image in a pair, computes the euclidean distance between these feature sets, and classifies the pair as either similar or dissimilar based on this distance.

Key Features

• Image Preprocessing: Load and normalize grayscale images for model input.
• Siamese Model Architecture: Build and compile a Siamese network with a custom distance layer to compare image features.
• Training and Validation: Train the model on a dataset of image pairs and validate its performance.
• Result Visualization: Plot training and validation loss and accuracy to evaluate the model's learning.

Installation

To set up the project environment, ensure you have Python installed, then run the following commands to install the necessary packages:

pip install tensorflow==2.3.0 opencv-python matplotlib

Prerequisites

• Python 3.8
• TensorFlow 2.3
• OpenCV
• Matplotlib
• Numpy

Setup

Clone this repository to your local machine:

git clone https://github.com/Rithik53/Siamese-Network-for-Image-Pair-Discrimination.git
cd siamese-image-pair-discrimination

Usage

To run this project, navigate to the project directory and execute the script by running:

python siamese_network.py

Example Output