12 lines (10 loc) · 436 Bytes

Contributing

Fork the repo and create a branch from master.
Install dependencies: pip install -r requirements.txt
Train the model: python train.py
Test your changes: python predict.py <your_file.wav>
Open a pull request.

Areas where help is especially welcome:

Unit tests for extract_features(), train(), and predict()
Improving model accuracy (currently ~65%)
Supporting additional audio formats