Skip to content

A deep learning classifier for urban sounds using the EfficientNet network

License

Notifications You must be signed in to change notification settings

katrin-ibrahim/Urban-Sound-Tagging-using-EfficientNet

Repository files navigation

Urban Sound Tagging using EfficientNet

The purpose of this project is to investigate how urban audio samples in computer-readable forms might be classified into the appropriate classes using an image transfer learning model. As a baseline, the EfficientNet-B1 model is used, along with Log-mel Spectrogram feature representations. To improve the baseline results,various audio data augmentation techniques are investigated. The two notebooks available are for the ESC-50/ESC-10 dataset and the UrbanSound8k dataset.

Best Results

  • 94.25% on ESC-10
  • 85.9% on ESC-50
  • 76.5% on UrbanSound8k

Authors

Acknowledgements

Feedback

If you have any feedback, please reach out to me at katrinemadfathy@gmail.com

To do

  • Experiment with larger network

  • Experiment with class conditional data augmentation

License

GPL v3

Releases

No releases published

Packages

No packages published