🎨 Neural Style Transfer with Van Gogh's Starry Night

This project applies Neural Style Transfer (NST) to blend the style of Vincent van Gogh’s Starry Night with the content of a custom image using deep learning. The implementation is done in Google Colab using TensorFlow and a pre-trained VGG19 model.

🧠 What is Neural Style Transfer?

Neural Style Transfer is a deep learning technique that combines two images:

• Content Image: The original image (e.g. a landscape or portrait).
• Style Image: The artwork whose painting style is transferred (here, Starry Night by Van Gogh).

By optimizing a loss function that measures content and style differences, a new image is generated that preserves the structure of the content image and mimics the style of the painting.

🖼️ Example Results

Content Image	Style Image (Starry Night)	Stylized Output

(You can replace these images with actual output examples from your notebook.)

🚀 Run It Yourself

Open the Colab notebook and run it step-by-step:

👉 Click here to open in Google Colab

🛠️ Technologies Used

• Python
• Google Colab
• TensorFlow / Keras
• VGG19 Pre-trained CNN
• Matplotlib, NumPy

⚙️ How It Works

1. Load and preprocess the content and style images.
2. Use VGG19 to extract features representing content and style.
3. Define a loss function combining content and style loss.
4. Optimize the image using gradient descent to minimize this loss.
5. Output the stylized image.

🧠 Learning Objectives

• Understand how convolutional layers extract visual features.
• Explore loss functions used for balancing content vs. style.
• Gain hands-on experience with TensorFlow and Colab.

🙋‍♂️ Author

Huseen Abo Ismael
4th-year Software Engineering Student