Cat vs Dog Classification using CNN

This project implements a Convolutional Neural Network (CNN) to classify images of cats and dogs. The model is trained on labeled images and predicts whether a given image contains a cat or a dog.

Features

• Uses a deep learning model (CNN) for image classification.
• Processes images of cats and dogs and predicts their category.
• Provides performance evaluation using Confusion Matrix, ROC, and precision-recall curves.

Dataset

The dataset consists of images of cats and dogs. You can use publicly available datasets such as:

• Kaggle Cats vs Dogs Dataset

Database Directory

As the dataset contains 25,000 images, they are not included in the repository. Below is the directory structure to organize the images after downloading from Kaggle:

| Cats-and-Dogs-Classification-using-CNN
|
| database
|        | 
|        | train
|        |     |
|        |     | dogs
|        |     |    |
|        |     |    |dog.1.jpg
|        |     | 
|        |     | cats
|        |     |    |
|        |     |    |cat.1.jpg
|        |
|        | test
|        |     |
|        |     | dogs
|        |     |    |
|        |     |    |dog.1.jpg
|        |     | 
|        |     | cats
|        |     |    |
|        |     |    |cat.1.jpg

Model Architecture

The CNN model consists of the following layers:

1. Convolutional Layers - Extracts features from images.
2. Pooling Layers - Reduces dimensionality.
3. Fully Connected Layers - Classifies images into cats or dogs.
4. Output Layer - Uses softmax/sigmoid for final classification.

Installation & Setup

1. Clone the repository:

   git clone https://github.com/mr-mahmood/Cats-and-Dogs-Classification-using-CNN.git
   cd Cats-and-Dogs-Classification-using-CNN

2. Install dependencies:

   pip install -r requirements.txt

Test with GUI

run the following file and test model:

python gui.py

Results

• The model achieves high accuracy in classifying cats and dogs around 95%.
• Performance metrics such as confusion matrix, precision, recall are used for evaluation.

Contact

For any questions or suggestions, feel free to reach out!