About Soundata

Soundata is a Python library for loading and working with audio datasets in a standardized way, removing the need for writing custom loaders in every project, and improving reproducibility by providing tools validate data against a canonical version. It speeds up research pipelines by allowing you to quickly download a dataset, load it into memory in a standardized and reproducible way, validate that the dataset is complete and correct, and more.

Soundata is based and inspired on mirdata, and was created following these desig principles:

  • Easy to use: Soundata is designed to be easy to use and to simplify the research pipeline considerably. Check out the examples in the Getting started page.

  • Easy to contribute to: we welcome and encourage contributions, especially new datasets. You can contribute following the instructions in our Contributing page.

  • Increase reproducibility: by providing a common framework for researchers to compare and validate their data, when mistakes are found in annotations or audio versions change, using Soundata the audio community can fix mistakes while still being able to compare methods moving forward.

  • Standardize usage of sound datasets: we standardized common attributes of sound datasets such as audio or tags to simplify audio research pipelines, while preserving each dataset’s idiosyncracies: if a dataset has ‘non-standard’ attributes, we include them as well.

Installation

To install Soundata simply do:

pip install soundata

We recommend to do this inside a conda or virtual environment for reproducibility.

Example Usage

import soundata

# learn wich datasets are available in soundata
print(soundata.list_datasets())

# choose a dataset and download it
dataset = soundata.initialize('urbansound8k', data_home='choose_where_data_live')
dataset.download()

# get annotations and audio for a random clip
example_clip = dataset.choice_clip()
tags = example_clip.tags
y, sr = example_clip.audio

Contributing

We welcome and encourage contributions to soundata, especially new dataset loaders. To contribute a new loader, please follow the steps indicated in the Contributing page and submit a Pull Request (PR) to the github repository. For any doubt or comment about your contribution, you can always submit an issue or open a discussion in the repository.

Citing soundata

If you refer to soundata’s design principles, motivation etc., please cite the following paper 1:

1

Fuentes, M., Salamon, J., Zinemanas, P., Rocamora, M., Plaja, G., Román, I.R., Miron, M., Serra, X. and Bello, J.P., 2021. Soundata: A Python library for reproducible use of audio datasets. arXiv e-prints, pp.arXiv-2109.

When working with datasets, please cite the version of soundata that you are using (given by the DOI above) AND include the reference of the dataset, which can be found in the respective dataset loader using the cite() method.