readme.md

# ArchIntelligent - Generative Design for Architects

**ArchIntelligent** is a deep learning project focused on **Generative AI for Architecture**. It enables architects and designers to quickly reimagine and redesign building concepts by generating new visuals from 3D building images (sketch/AutoCAD/Sketchup). Whether it's exploring new styles, adjusting functional details, or applying custom prompts, this tool accelerates the creative process. A simple, interactive **Streamlit app** is also included to provide a user-friendly interface for experimenting with the model.

## 🌐 Try It Online (Hugging Face Space)
👉 [Launch ArchIntelligent on Hugging Face Spaces](https://huggingface.co/spaces/harrydawitch/ArchIntelligent)

No setup needed. Just upload your 3D building image, select your style and prompt, and generate your design.

>⚠️ Note: Due to my financial limitations, my Hugging Face Space currently runs on CPU, which makes image generation quite slow (around 30 minutes per image).
For a faster experience, consider cloning this repo and running it locally on a machine with a GPU.

## 🚀 Features
- 🎨 Style transfer for architectural designs
- 🏢 Generate redesigned images from 3D exterior building image inputs (interior coming soon)
- 💡 Customize design output with prompts, style, functional, environment and creative scale tweaks
- 🌐 Streamlit web interface for easy interaction

## 🧪 Running Locally
**1. Clone the repo:**
```console
git clone https://github.com/harrydawitch/ArchIntelligent.git
cd ArchIntelligent
```

**2. Install dependencies:**
```console
pip install -r requirements.txt
```

**3. Launch the app:**

```console
streamlit run app.py
```

## 📸 Example
**Here’s a sample workflow:**
1. Upload your 3D building render
2. Choose a design style (e.g. Modern, Minimal, Gothic, Futuristic...)
3. Choose building functional(e.g. Villa, Office, Apartment, Hotel)
4. Describe your design using prompts (e.g."Exterior view of a office tower with glass facade, green rooftop" etc.)
5. Generate!
   
## 🤝 Contributing
Pull requests are welcome! For major changes, please open an issue first to discuss what you'd like to change or improve.


## 📄 License
MIT License – see the LICENSE file for details.