Sleep analysis from sound recording using Pytorch. First aim of this project is to build and tune general sound classifiers to detect automatically somniloquy. The current most effective implemented approach is based on CNN applied directly to the sound spectrogram images.
Run the following command to populate your data/ folder with the datasets:
# fetch the datasets from Google Drive
poetry run python dev/get_data.pyThe datasets from your local data/ folder will be automatically updated to Google
Drive as a .zip (reserved to maintainers):
# upload the local data/ folder to Google Drive
poetry run python dev/upload_data_to_drive.pyIf you wish to contribute to this project please head to CONTRIBUTING.md .