# 🚀 Quick Start Guide

## One-Command Setup

### Method 1: Using our setup script
```bash
# Download the setup script
curl -O https://huggingface.co/zhiqing0205/u2net-mvtec-loco-segmentation/raw/main/setup_project.py

# Run setup (downloads everything automatically)
python setup_project.py

# Use the project
cd u2net-mvtec-loco
python mvtec_loco_fg_segmentation.py
```

### Method 2: Using HuggingFace CLI
```bash
# Install HuggingFace CLI
pip install huggingface_hub

# Download complete project (equivalent to git clone)
huggingface-cli download zhiqing0205/u2net-mvtec-loco-segmentation \
    --local-dir ./u2net-project --repo-type model

# Use the project
cd u2net-project
python mvtec_loco_fg_segmentation.py
```

### Method 3: Using Python
```bash
# One-liner to download everything
python -c "
from huggingface_hub import snapshot_download
snapshot_download('zhiqing0205/u2net-mvtec-loco-segmentation', local_dir='./u2net-project')
print('Done! cd u2net-project && python mvtec_loco_fg_segmentation.py')
"
```

## What Gets Downloaded

✅ Complete source code  
✅ Pre-trained model weights (u2net.pth - 169MB)  
✅ Documentation (English + Chinese)  
✅ Example scripts and utilities  
✅ Ready to run immediately  

## File Structure After Download
```
u2net-mvtec-loco/
├── mvtec_loco_fg_segmentation.py  # Main script
├── saved_models/
│   └── u2net/
│       └── u2net.pth             # Pre-trained model (169MB)
├── model/                         # Model architecture
├── data_loader.py                 # Data utilities
├── README.md                      # English docs
├── README_CN.md                   # Chinese docs
└── ...
```

## Immediate Usage
```bash
# Process entire MVTec LOCO dataset
python mvtec_loco_fg_segmentation.py

# Process specific categories
python mvtec_loco_fg_segmentation.py --categories breakfast_box

# Custom threshold
python mvtec_loco_fg_segmentation.py --threshold 0.3
```

That's it! 🎉