🎙️ Bilingual Audio Recorder (English & Tulu)

An interactive Python tool to record bilingual speech samples using PyAudio and ipywidgets in a Jupyter Notebook. Ideal for collecting aligned audio datasets (e.g., English ↔ Tulu) for speech recognition, language translation, or linguistic research.

---

📦 Features

• ✅ Record audio for two separate languages (English and Tulu)
• 🧠 Interactive UI using ipywidgets for Start/Stop per language
• 💾 Automatically stores WAV files with a unique random ID
• 📝 Input label field and CSV logging (data.csv)
• ❌ Cancel button to discard unwanted recordings
• 📂 Organizes files into separate folders (ENGLISH/, TULU/)

---

🛠️ Requirements

• Python 3.7+
• pyaudio
• ipywidgets
• Jupyter Notebook or JupyterLab

Install dependencies

pip install pyaudio ipywidgets

For Linux, if PyAudio fails to install:

sudo apt-get install portaudio19-dev
pip install pyaudio

---

🚀 How to Use

1. Launch your Jupyter Notebook:

   jupyter notebook

2. Open the notebook containing the code.

3. Click Start Recording ENGLISH and Start Recording TULU as needed.

4. Click Stop when done.

5. Enter the associated label/text in the input box.

6. Press NEXT to save the label in data.csv.

7. If you want to discard the current recording pair, press Cancel.

Files will be saved as:

ENGLISH/<randomID>.wav
TULU/<randomID>.wav

And your CSV (data.csv) will contain:

<randomID>,<user_input_text>

---

🧩 Project Structure

bilingual-audio-recorder/
├── ENGLISH/               # Saved English audio files
├── TULU/                  # Saved Tulu audio files
├── data.csv               # Text labels for each recording pair
├── notebook.ipynb         # Main notebook (contains all UI and logic)
└── README.md

---

🤝 Contributing

Contributions are welcome! If you'd like to:

• Add support for more languages
• Include audio visualizations
• Improve file organization Feel free to open an issue or submit a pull request.

---

📜 License

MIT License