Real-ESRGAN Mac GUI

A simple, user-friendly graphical interface for Real-ESRGAN on macOS. This GUI provides an easy way to upscale images using Real-ESRGAN's powerful AI models without touching the command line.

Features

• Drag-and-drop support for easy file management
• Multiple AI models (RealESRGAN_x4plus, anime models, and more)
• Customizable upscale ratio (1x to 8x)
• Optional face enhancement for portrait images
• Batch processing support for multiple images
• Real-time log output to monitor progress
• Simple, clean interface built with PySide6

Prerequisites

Required Software

1. macOS (tested on macOS 10.15+)
2. Python 3.8 or higher
3. Real-ESRGAN repository - You need a local clone of the official Real-ESRGAN repository

Setting Up Real-ESRGAN

If you haven't already set up Real-ESRGAN, follow these steps:

# Clone the Real-ESRGAN repository
git clone https://github.com/xinntao/Real-ESRGAN.git
cd Real-ESRGAN

# Install Real-ESRGAN dependencies
pip3 install basicsr
pip3 install facexlib
pip3 install gfpgan
pip3 install -r requirements.txt
python setup.py develop

# Download pre-trained models
wget https://github.com/xinntao/Real-ESRGAN/releases/download/v0.1.0/RealESRGAN_x4plus.pth -P weights
wget https://github.com/xinntao/Real-ESRGAN/releases/download/v0.2.2.4/RealESRGAN_x4plus_anime_6B.pth -P weights

Installation

1. Download the GUI script:

   curl -O https://raw.githubusercontent.com/[your-repo]/realesrgan_gui.py

2. Install GUI dependencies:

   pip3 install PySide6

Usage

Starting the Application

python3 realesrgan_gui.py

First-Time Setup

1. Click "Set inference_realesrgan.py" button
2. Navigate to your Real-ESRGAN clone directory
3. Select the inference_realesrgan.py file (usually at ~/Real-ESRGAN/inference_realesrgan.py)

Processing Images

1. Add images:

   • Click "Add File/Folder" button, OR
   • Drag and drop image files directly onto the file list

2. Configure settings:

   • Model: Choose the AI model based on your content type
     • RealESRGAN_x4plus - General photos and images
     • RealESRGAN_x4plus_anime_6B - Anime and illustrations
     • RealESRNet_x4plus - More conservative upscaling
   • Outscale: Set the upscaling ratio (1-8x)
   • Face enhance: Check this for better results on portraits

3. Set output location (optional):

   • Click "Choose Output Folder" to specify where upscaled images will be saved
   • Default: ./results in the current directory

4. Start processing:

   • Click "Start" to begin upscaling
   • Monitor progress in the log window
   • Use "Stop" button to cancel if needed

Tips for Best Results

• For photos: Use RealESRGAN_x4plus with face enhancement enabled for portraits
• For anime/drawings: Use RealESRGAN_x4plus_anime_6B for sharper, cleaner results
• Upscale ratio: Start with 4x for most cases; higher ratios work best with very low-resolution images
• Batch processing: You can add multiple images; the GUI will process them sequentially

Supported Models

Model	Best For	File Size
RealESRGAN_x4plus	General images, photos	~65MB
RealESRGAN_x4plus_anime_6B	Anime, illustrations, art	~17MB
RealESRNet_x4plus	Conservative upscaling	~65MB

Troubleshooting

"Script not found" error

• Make sure you've set the correct path to inference_realesrgan.py
• Verify the Real-ESRGAN repository is properly installed
• Check that the path doesn't contain spaces or special characters

"No module named..." errors

• Ensure all Real-ESRGAN dependencies are installed
• Try reinstalling: pip3 install basicsr facexlib gfpgan

GUI doesn't start

• Verify Python version: python3 --version (must be 3.8+)
• Reinstall PySide6: pip3 install --upgrade PySide6

Slow processing

• GPU acceleration is automatically used if available
• Processing time depends on image size and upscale ratio
• Larger images and higher ratios take longer

File Structure

.
├── realesrgan_gui.py    # This GUI application
├── results/             # Default output folder (created automatically)
└── Real-ESRGAN/         # Your Real-ESRGAN clone (required)
    ├── inference_realesrgan.py
    ├── weights/
    └── ...

Technical Details

• GUI Framework: PySide6 (Qt for Python)
• Backend: Calls Real-ESRGAN's inference_realesrgan.py as a subprocess
• Threading: Uses worker threads to prevent UI freezing during processing
• Logging: Real-time output capture from the Real-ESRGAN process

Limitations

• Currently processes one image at a time (sequential batch processing)
• Requires a local Real-ESRGAN installation
• No built-in model downloader (manual setup required)
• macOS only (Linux/Windows would need minor modifications)

Contributing

Contributions are welcome! Feel free to:

• Report bugs
• Suggest new features
• Submit pull requests

License

This GUI wrapper is provided as-is. Please refer to the Real-ESRGAN license for the underlying AI models and processing code.