Online Speech Translator

A real-time speech-to-text and translation tool with a live terminal interface.

Built on top of the excellent KoljaB/RealtimeSTT library.

Features

• Real-time speech recognition from microphone (with available input device listing)
• Asynchronous translation using deep-translator (Google Translate) or OpenAI
• Live updating table in the terminal using rich
• Device selection and device listing via command line
• Proxy support for translation requests
• Language selection for both input and translation
• Support for multiple languages
• Logging of transcripts and translations (see app.log, transcript.log, transcript_with_translation.log)
• Context management for improved translation accuracy
• Modular translation and rendering backends (see translator/ and renderer/)

Project Structure

├── realtime_stt.py                  # Main logic for real-time speech translation
├── input_devices.py                 # Audio device utilities
├── translator/                      # Translation interfaces and implementations
│   ├── __init__.py
│   ├── base.py
│   ├── factory.py
│   ├── google_translator.py
│   └── openai_translator.py
├── compressor/                      # Context compressor for OpenAI
│   ├── __init__.py
│   ├── base.py
│   └── openai_compressor.py
├── renderer/                        # Output rendering (terminal, HTML, etc.)
│   ├── __init__.py
│   ├── base.py
│   ├── factory.py
│   ├── html_fastaip_renderer.py
│   └── rich_render.py
├── requirements.txt                 # Project dependencies
├── readme.md                        # Documentation
├── app.log                          # Application log
├── transcript.log                   # Raw transcript log
├── transcript_with_translation.log  # Transcript with translation log
├── LICENSE                          # License file

Quickstart

1. Create a Python Environment

Option A: Using Conda

conda create -n online_speech_translate python=3.12
conda activate online_speech_translate

Option B: Using venv

python3 -m venv .venv
source .venv/bin/activate  # On Windows use: .venv\Scripts\activate

2. Install Dependencies

Linux Installation

Before installing dependencies, run:

sudo apt-get update
sudo apt-get install python3-dev portaudio19-dev

MacOS Installation

Before installing dependencies, run:

brew install portaudio

Then install Python dependencies:

git clone https://github.com/nikkiw/realtime_translator.git
cd realtime_translator
pip install -r requirements.txt

3. List Available Audio Devices

python realtime_stt.py --list_devices

4. Run the Application

To run the real-time speech translator, execute:

python realtime_stt.py --input_device_index <device_index> --input_lang <source_language> --translate_lang <target_language>

Command Line Arguments

• --input_device_index: Index of the audio input device (default is 0).
• --input_lang: Language spoken by the speaker (2-letter code, e.g., 'en', one of: en, ru, de, pt, it, es, fr).
• --translate_lang: Target translation language (2-letter code, e.g., 'ru', one of: en, ru, de, pt, it, es, fr).
• --proxy: Optional proxy URL for translation requests.
• --translator_type: Type of translator to use ('google' or 'openai').
• --openai_api_key: OpenAI API key for using the OpenAI translator.
• --renderer: Output rendering engine. Options:
  • rich: Shows a live-updating, color table in the terminal (recommended for CLI use).
  • html_fastaip: Outputs results to an HTML page (convenient for web integration or browser viewing, url: http://127.0.0.1:8090).
• --log_level: Logging level (default: INFO).
• --list_devices: List available audio devices and exit.

Example with Proxy and HTML Renderer

python realtime_stt.py --input_device_index 0 --input_lang en --translate_lang ru --renderer html_fastaip --proxy http://user:pass@host:port

Renderer Options

• rich: Displays results in a modern, colorized table directly in your terminal. Supports live updates and is ideal for command-line workflows.
• html_fastaip: Renders results to an HTML file for viewing in a browser or embedding in web apps. Useful for sharing or integrating with other tools.

Requirements

• Python 3.12
• See requirements.txt for Python dependencies

Logs

• app.log: General application log
• transcript.log: Raw recognized text
• transcript_with_translation.log: Recognized text with translation

Contributing

Contributions are welcome! Please submit a pull request or open an issue for any enhancements or bug fixes.

License

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

---