README.md

metadata

title: Demucs Music Source Separation
emoji: 🎵
colorFrom: purple
colorTo: pink
sdk: docker
pinned: false
license: mit

Demucs Music Source Separation

An AI-powered web application for separating music into individual stems (vocals, drums, bass, other) using Meta's Demucs model.

Features

• 🎵 High-Quality Separation: Uses the latest Demucs htdemucs model for superior audio separation
• 🎨 Beautiful UI: Modern React interface with real-time progress tracking
• 📱 Responsive Design: Works on desktop and mobile devices
• ⚡ Fast Processing: Optimized for quick audio processing
• 💾 Easy Download: Download individual stems as WAV files

Supported Audio Formats

• MP3
• WAV
• FLAC
• OGG
• M4A
• AAC

How to Use

1. Upload your audio file using the drag-and-drop interface
2. Wait for the AI to process and separate the audio (progress bar shows status)
3. Listen to individual stems using the built-in audio players
4. Download the separated tracks you want to keep

Technology Stack

• Frontend: React + Vite
• Backend: Flask + Python
• AI Model: Meta's Demucs (Hybrid Transformer)
• Audio Processing: PyTorch + TorchAudio
• Deployment: Docker + Hugging Face Spaces

Local Development

# Install dependencies
npm install
pip install -r requirements.txt

# Start development servers
npm run dev          # Frontend (port 3001)
python server.py     # Backend (port 5000)

Credits

• Demucs by Meta Research
• Built with ❤️ for the music community

License

MIT License - feel free to use and modify!