Dataset used in COALA: Co-Aligned Autoencoders for Learning Semantically Enriched Audio Representations

<p>This dataset consists of two hdf5 files that contain pre-computed log-mel spectrograms that have been used to to train audio embedding models. The dataset is split into a training set and a validation set containing respectively 170793 and&nbsp;19103 spectrogram patches with their accompanying multi-hot encoded tags from a vocabulary of 1000 tags provided by <a href="https://freesound.org/">Freesound</a> users.</p> <p>More details can be found in &quot;COALA: Co-Aligned Autoencoders for Learning Semantically Enriched Audio Representations&quot; by X. Favory, <a href="https://kdrossos.net">K. Drossos</a>, <a href="https://tutcris.tut.fi/portal/en/persons/tuomas-virtanen(210e58bb-c224-40a9-bf6c-5b786297e841).html">T. Virtanen</a>, and X. Serra. The code is available at this <a href="https://github.com/xavierfav/coala">GitHub repository</a>.</p> <p>&nbsp;</p> <p>License:</p> <p>This dataset is derived from content from the Freesound collection. All sounds are released under Creative Commons (CC) licenses from either <a href="https://creativecommons.org/publicdomain/zero/1.0/">CC0</a>, <a href="https://creativecommons.org/licenses/by/3.0/">CC-BY,</a> <a href="https://creativecommons.org/licenses/sampling+/1.0/">CC-S+</a>, or <a href="https://creativecommons.org/licenses/by-nc/3.0/">CC-BY-NC</a>. We attribute authors of all the sounds used in the dataset and provide their corresponding licenses in the attributions.txt file.</p> <p>&nbsp;</p>