PreTrained Models for Image Classification

This repository contains implementations of various deep learning models for image classification tasks, including Convolutional Neural Networks (CNNs) and Transfer Learning models.

Contents

1. LeNet-5 Pretrained Model:

   • Implementation of LeNet-5 pretrained model on MNIST and CIFAR-10 datasets.
   • Training and evaluation of the model.
   • Visualization of the model architecture, training history, confusion matrix, and classification report.

2. ResNet-50 Transfer Learning:

   • Implementation of transfer learning with the ResNet-50 model on MNIST and CIFAR-10 datasets.
   • Training and evaluation of the model.
   • Visualization of the model architecture, training history, confusion matrix, and classification report.

3. Inception Model:

   • Implementation of an Inception model for MNIST classification.
   • Training and evaluation of the model.
   • Visualization of the model architecture, training history, confusion matrix, and classification report.

4. Xception Transfer Learning:

   • Implementation of transfer learning with the Xception model on MNIST dataset.
   • Hyperparameter optimization using Optuna.
   • Training and evaluation of the model.
   • Training and evaluation of Random Forest Classifier on extracted features.
   • Visualization of precision, recall, F1 score, and confusion matrix.

Prerequisites

Ensure you have the following libraries installed:

• TensorFlow
• NumPy
• Matplotlib
• Seaborn
• Scikit-learn
• Optuna (for Xception Transfer Learning)

Usage

To use the code:

1. Clone the repository:

   git clone <repository_url>

2. Navigate to the repository directory:

   cd <repository_directory>

3. Run the respective Python scripts to train and evaluate the models for image classification.

Author

• Mohamed Abobakr
• Nour Hamdy
• Mariam Elsabagh