🎨 Neural Style Transfer Web Application

A Flask-based web application that applies artistic styles to images using Neural Style Transfer (NST) implemented with PyTorch.

🚀 Features

• Upload a Content Image
• Upload a Style Image
• Generate artistic stylized images
• Deep Learning-based Neural Style Transfer
• Interactive Flask Web Interface
• Real-time image processing

🛠️ Tech Stack

• Python
• Flask
• PyTorch
• TorchVision
• HTML
• CSS
• Bootstrap
• Pillow

📂 Project Structure

NST_Code/
│
├── app.py
├── train.py
├── requirements.txt
├── vgg_normalised.pth
│
├── Content_data/
├── Style_data/
├── experiment/
│
├── templates/
│   └── index.html
│
├── static/
│   ├── uploads/
│   ├── css/
│   └── images/
│
└── utils/
    ├── models.py
    └── utils.py

⚙️ Installation

Clone Repository

git clone https://github.com/siddemmohankrishna/Neural-Style-Transfer-WebApp.git
cd Neural-Style-Transfer-WebApp

Create Virtual Environment

python -m venv myenv

Activate Environment

Windows:

myenv\Scripts\activate

Linux/Mac:

source myenv/bin/activate

Install Dependencies

pip install -r requirements.txt

🏋️ Model Training

python train.py --batch_size 4 --epochs 10 --experiment trial2

Training checkpoints will be saved inside:

experiment/trial2/

🌐 Run Web Application

python app.py

Open:

http://127.0.0.1:5000

in your browser.

📸 Example Workflow

1. Upload a Content Image
2. Upload a Style Image
3. Click Generate
4. Download the Stylized Output

📈 Deep Learning Architecture

• Encoder: Pre-trained VGG Network
• Adaptive Instance Normalization (AdaIN)
• Decoder Network
• Content Loss
• Style Loss

🎯 Future Improvements

• Multiple style blending
• GPU acceleration
• User authentication
• Cloud deployment
• Batch image processing

👨‍💻 Author

Siddem Mohan Krishna

• GitHub: https://github.com/siddemmohankrishna
• LinkedIn: https://www.linkedin.com/in/siddem-mohan-krishna-247984378/

📜 License

This project is developed for educational and research purposes.