German Live Translator

A simple live translation app that transcribes German audio to text using a local model (Faster-Whisper) and provides a word-by-word English translation using a local dictionary.

Prerequisites

• Python 3.8+
• Node.js 16+
• ffmpeg (required for faster-whisper audio processing)
  • Mac: brew install ffmpeg
  • Linux: sudo apt install ffmpeg
  • Windows: Download and add to PATH

Setup

1. Backend Setup

   cd backend
   python3 -m pip install -r requirements.txt

2. Frontend Setup

   cd frontend
   npm install

Running the App

1. Start the Backend

   cd backend
   python3 main.py

   The server will start at http://0.0.0.0:8000.

2. Start the Frontend

   cd frontend
   npm run dev

   Open the URL shown (usually http://localhost:5173) in your browser.

Usage

1. Allow microphone access when prompted.
2. Click the Mic button to start recording.
3. Speak in German.
4. The app will transcribe your speech and show word-by-word translations below the German text.
5. Click the Stop button to stop recording.

Notes

• The first time you run the backend, it will download the faster-whisper model (approx. 500MB - 1GB depending on the model size).
• The "live" translation works by sending audio chunks every 3 seconds.
• The dictionary is loaded from backend/dictionary.json.