🖼️ Neural Style Transfer

This project implements Neural Style Transfer (NST) using Adaptive Instance Normalization (AdaIN).
The encoder is based on VGG, and the decoder reconstructs the stylized image.

---

🚀 Features

• Upload a content image and a style image.
• Generates a new image that preserves the content of the first image while adopting the style of the second.
• Uses AdaIN for flexible style transfer.
• Encoder: VGG network pretrained on ImageNet.

---

📂 How It Works

1. Content image → passed through VGG encoder.
2. Style image → also passed through VGG encoder.
3. AdaIN aligns the statistics (mean & variance) of content features to style features.
4. Decoder reconstructs the stylized image.

---

Glimpses from app