ReFaceX: AI-Powered Face Swapping & Enhancement Tool

ReFaceX is a powerful AI-driven tool that allows you to seamlessly swap faces and enhance image quality with just one click! It leverages the latest advancements in face enhancement and swapping technologies, providing professional-level results in seconds. Whether you're working in film, advertising, or content creation, ReFaceX delivers both precision and speed.

📌 Examples

Input 1	Input 2	Swapped Ouput

✨ Features

• 🎭 AI-Powered Face Enhancement
  Using state-of-the-art models like GFPGAN and Real-ESRGAN, ReFaceX significantly improves the quality of images, restoring fine facial details.

• 🔄 Realistic Face Swapping
  Powered by InsightFace, our face-swapping technology is precise and natural, giving lifelike results that are indistinguishable from reality.

• ⚡ Fast & Efficient
  Optimized for speed without sacrificing accuracy, making it easy for users to generate high-quality results in minimal time.

• 💼 Professional-Grade Output
  Tailored for professionals in film, TV, advertising, and design, but also accessible for casual users.

🛠️ Tech Stack

• GFPGAN & Real-ESRGAN – For facial enhancement and image super-resolution.
• InsightFace – For accurate face detection, alignment, and swapping.
• Python – Core programming language for AI model integration.
• Flask – Backend framework for handling model operations.
• Streamlit – For the interactive web interface (optional).

📂 Directory Structure

└── ReFaceX/
    ├── face_enhancer.py        # Face enhancement module
    ├── faceswap.py             # Face swapping module
    ├── main.py                 # Main application script
    ├── requirements.txt        # Python dependencies
    ├── images/                 # Sample/test images
    │   └── Image5.avif
    ├── model/                  # Pretrained models
    │   ├── FaceFusion-SoC.onnx
    │   ├── PreTrainedRealESRGAN_x2.pth
    │   └── RealESRGAN_x4.pth
    ├── streamlit_app/          # Streamlit web app (optional)
    │   ├── __init__.py
    │   ├── app.py
    │   ├── utils.py
    │   └── styles/
    │       └── custom.css
    └── upscaler/               # RealESRGAN upscaling module
        ├── __init__.py
        └── RealESRGAN/
            ├── __init__.py
            ├── arch_utils.py
            ├── model.py
            ├── rrdbnet_arch.py
            └── utils.py

🚀 Quick Start

Prerequisites

• Python 3.8+
• CUDA (for GPU acceleration, optional)
• pip

Installation

1. Clone the repository:

   git clone https://github.com/vineetg2003/ReFaceX.git
   cd ReFaceX

2. Install dependencies:

   pip install -r requirements.txt

3. Download required models and place them in the model/ directory.

Usage

Command Line Interface (CLI)

Run the main script:

python main.py --input input.jpg --output output.jpg --mode [enhance|swap]

Web Interface (Streamlit)

If using the Streamlit app:

streamlit run streamlit_app/app.py

📄 License

This project is licensed under the MIT License. See LICENSE for details.

🤝 Contributing

Contributions are welcome! Please open an issue or submit a pull request.

📧 Contact

For inquiries, contact:
📩 vineetg2003@gmail.com
🌐 https://github.com/vineetg2003