Cat vs Dog Image Classifier 🐱🐶

A Convolutional Neural Network (CNN) built with TensorFlow/Keras to classify images of cats and dogs with high accuracy.

📋 Project Overview

This project implements a binary image classification model that can distinguish between cats and dogs. The CNN architecture includes multiple convolutional layers, pooling layers, and dense layers to effectively learn features from pet images.

🏗️ Model Architecture

Sequential Model:
├── Conv2D (32 filters, 3x3 kernel, ReLU)
├── MaxPooling2D (2x2)
├── Conv2D (32 filters, 3x3 kernel, ReLU) 
├── MaxPooling2D (2x2)
├── Flatten
├── Dense (128 units, ReLU)
└── Dense (1 unit, Sigmoid)

📊 Dataset Information

• Training Set: 8,000 images (4,000 cats + 4,000 dogs)
• Test Set: 2,000 images (1,000 cats + 1,000 dogs)
• Image Size: 64x64 pixels
• Format: RGB images

🚀 Features

• Data Augmentation: Rescaling, shearing, zooming, and horizontal flipping
• Binary Classification: Outputs probability for cat (0) or dog (1)
• Real-time Prediction: Can classify new images instantly
• Optimized Architecture: Efficient CNN design for quick training

🛠️ Installation & Setup

Prerequisites

Python 3.7+
TensorFlow 2.x
NumPy
Keras

Install Dependencies

pip install tensorflow numpy keras pillow

Clone Repository

git clone https://github.com/Ahmadhammam03/cat-dog-classifier.git
cd cat-dog-classifier

📁 Project Structure

cat-dog-classifier/
│
├── cnn.ipynb                 # Main Jupyter notebook
├── dataset/
│   ├── training_set/
│   │   ├── cats/            # 4,000 cat images
│   │   └── dogs/            # 4,000 dog images
│   ├── test_set/
│   │   ├── cats/            # 1,000 cat images
│   │   └── dogs/            # 1,000 dog images
│   └── single_prediction/
│       ├── cat_or_dog_1.jpg # Sample cat image
│       └── cat_or_dog_2.jpg # Sample dog image
├── models/
│   └── saved_model/         # Trained model files
├── README.md
└── requirements.txt

🎯 Usage

Training the Model

# Run the Jupyter notebook
jupyter notebook cnn.ipynb

# Or run as Python script
python cnn.py

Making Predictions

import numpy as np
from tensorflow.keras.preprocessing import image
from tensorflow.keras.models import load_model

# Load trained model
model = load_model('models/cat_dog_classifier.h5')

# Load and preprocess image
test_image = image.load_img('path/to/image.jpg', target_size=(64, 64))
test_image = image.img_to_array(test_image)
test_image = np.expand_dims(test_image, axis=0)

# Make prediction
result = model.predict(test_image)
prediction = 'dog' if result[0][0] > 0.5 else 'cat'
print(f"Prediction: {prediction}")

📈 Model Performance

• Training Accuracy: ~95%
• Validation Accuracy: ~90%
• Epochs: 25
• Optimizer: Adam
• Loss Function: Binary Crossentropy

🔧 Hyperparameters

Parameter	Value
Batch Size	32
Image Size	64x64
Epochs	25
Learning Rate	Default (Adam)
Filters (Conv1)	32
Filters (Conv2)	32
Dense Units	128

🎨 Data Augmentation

The model uses several augmentation techniques to improve generalization:

• Rescaling: Normalizes pixel values (1./255)
• Shear Range: 0.2
• Zoom Range: 0.2
• Horizontal Flip: Random horizontal flipping

🚧 Future Improvements

• Implement transfer learning with pre-trained models (VGG16, ResNet)
• Add more animal classes (birds, rabbits, etc.)
• Increase image resolution for better accuracy
• Deploy model as web application
• Add model visualization and interpretability features

🤝 Contributing

1. Fork the repository
2. Create your feature branch (git checkout -b feature/AmazingFeature)
3. Commit your changes (git commit -m 'Add some AmazingFeature')
4. Push to the branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

👨‍💻 Author

• GitHub: Ahmadhammam03
• LinkedIn: Ahmad Hammam

🙏 Acknowledgments

• TensorFlow/Keras documentation
• Deep Learning community
• Dataset providers

---

⭐ If you found this project helpful, please give it a star! ⭐