Audio Separator Web Interface

This is a web interface for the Mel-Band-Roformer Vocal Model that separates vocals from music tracks.

Features

• Simple web interface
• Supports multiple audio formats (MP3, WAV, OGG, FLAC, M4A)
• Automatic model and configuration download
• Real-time processing status updates

Installation

1. Clone the repository:

git clone https://github.com/Atm4x/Mel-Band-Roformer-Vocal-Model-GUI
cd audio-separator

2. Create and activate a virtual environment (optional but recommended):

python -m venv venv
source venv/bin/activate  # On Windows use: venv\Scripts\activate

3. Install requirements:

pip install -r requirements.txt

Running the Application

Simply run:

python inference.py

The application will:

1. Automatically download the required model and configuration files
2. Start a web server at http://localhost:5000

Usage

1. Open your web browser and go to http://localhost:5000
2. Upload an audio file using the web interface
3. Wait for processing to complete
4. Download the separated vocal and instrumental tracks

Notes

• The first run will download the model (~900MB) and configuration files
• Processing time depends on your computer's specifications and the length of the audio file
• Generated files are stored in the outputs folder

Requirements

• Python 3.7 or higher
• At least 6GB RAM
• GPU is recommended but not required

License

This project is licensed under the MIT License - see the LICENSE file for details.