🐶🐱 Dog vs Cat Image Classification using Deep Learning

This project implements a Convolutional Neural Network (CNN) using TensorFlow/Keras to classify images of dogs and cats. It follows a clean dataset structure, professional project layout, and is suitable for academic projects, internships, and GitHub portfolios.

---

📂 Project Structure

dog-cat-classifier/
│
├── dataset/
│   ├── train/
│   │   ├── cats/
│   │   └── dogs/
│   │
│   ├── validation/
│       ├── cats/
│       └── dogs/
│
├── cat.jpg                      # Sample cat image
├── dog.jpg                      # Sample dog image
├── dog_cat_classifier.keras     # Trained Keras model
├── Figure_1.png                 # Training accuracy/loss graph
├── Figure_2.png                 # Validation accuracy/loss graph
├── train.py                     # Model training script
├── test.py                      # Model testing / inference script
├── requirements.txt             # Project dependencies
├── README.md                    # Project documentation

---

🚀 Features

• Binary image classification (Dog vs Cat)
• CNN built using TensorFlow & Keras
• Organized dataset structure (train / validation)
• Model evaluation with accuracy graphs
• Trained model saved in native Keras format
• Easy to run on local system

---

🧠 Technologies Used

• Python 3.9
• TensorFlow / Keras
• NumPy
• Matplotlib
• Pillow (PIL)
• SciPy

---

📊 Dataset Structure

The dataset follows the standard Keras directory format:

dataset/
├── train/
│   ├── cats/
│   └── dogs/
│
├── validation/
    ├── cats/
    └── dogs/

• Images in cats/ → Label 0
• Images in dogs/ → Label 1

---

⚙️ Installation & Setup

Model link

https://www.kaggle.com/models/dharshiyanacc/dog-cat-classifier

Dataset link

https://www.kaggle.com/datasets/dharshiyanacc/dog-vs-cat

1️⃣ Clone the repository

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

2️⃣ Create virtual environment (recommended)

python -m venv venv
venv\Scripts\activate   # Windows

3️⃣ Install dependencies

pip install -r requirements.txt

---

▶️ Training the Model

Run the training script:

python train.py

• Trains the CNN model

• Saves the trained model as:

  dog_cat_classifier.keras
  

• Generates training & validation accuracy plots

---

🧪 Testing / Prediction

Run the test script:

python test.py

Use sample images like:

• cat.jpg
• dog.jpg

---

📈 Results

• Training and validation accuracy graphs are saved as:

  • Figure_1.png
  • Figure_2.png

• Model achieves reliable classification performance on validation data

---

💾 Model Format

The trained model is saved using native Keras format:

dog_cat_classifier.keras

Why .keras format?

• Recommended by TensorFlow
• Preserves model architecture + weights
• Better forward compatibility than .h5

---

📌 Future Improvements

• Add test dataset
• Data augmentation
• Transfer learning (MobileNet / ResNet)
• Convert model to TensorFlow Lite
• Build a web interface for predictions

---

👨‍💻 Author

DHARSHIYAN - Deep Learning & Computer Vision Enthusiast

---

📄 License

This project is licensed under the MIT License.