MUSDB18 - a corpus for music separation

<p>The sigsep musdb18 data set consists of a total of 150 full-track songs of different styles and includes both the stereo mixtures and the original sources, divided between a training subset and a test subset.</p> <p>Its purpose is to serve as a reference database for the design and the evaluation of source separation algorithms. The objective of such signal processing methods is to estimate one or more sources from a set of mixtures, e.g. for karaoke applications. It has been used as the official dataset in the professionally-produced music recordings task for SiSEC 2018, which is the international campaign for the evaluation of source separation algorithms.</p> <p><em>musdb18</em> contains two folders, a folder with a training set: &ldquo;train&rdquo;, composed of 100 songs, and a folder with a test set: &ldquo;test&rdquo;, composed of 50 songs. Supervised approaches should be trained on the training set and tested on both sets.</p> <p>All files from the <em>musdb18</em> dataset are encoded in the Native Instruments stems format (.mp4). It is a multitrack format composed of 5 stereo streams, each one encoded in AAC @256kbps. These signals correspond to:</p> <ul> <li>0 - The mixture,</li> <li>1 - The drums,</li> <li>2 - The bass,</li> <li>3 - The rest of the accompaniment,</li> <li>4 - The vocals.</li> </ul> <p>For each file, the mixture correspond to the sum of all the signals. All signals are stereophonic and encoded at 44.1kHz.</p> <p>As the <em>MUSDB18</em> is encoded as STEMS, it relies on ffmpeg to read the multi-stream files. We provide a python wrapper called stempeg that allows to easily parse the dataset and decode the stem tracks on-the-fly.</p> <p><strong>License</strong></p> <p>MUSDB18 is provided for educational purposes only and the material contained in them should not be used for any commercial purpose without the express permission of the copyright holders:</p> <ul> <li>100 tracks were derived from The &lsquo;Mixing Secrets&rsquo; Free Multitrack Download Library. Please refer to this original resource for any question regarding your rights on your use of the DSD100 data.</li> <li>46 tracks are taken from the MedleyDB licensed under Creative Commons (BY-NC-SA 4.0).</li> <li>2 tracks were kindly provided by Native Instruments originally part of their stems pack.</li> <li>2 tracks a from from the Canadian rock band The Easton Ellises as part of the heise stems remix competition, licensed under Creative Commons (BY-NC-SA 3.0).</li> </ul> <p><strong>References</strong></p> <p>If you use the MUSDB dataset for your research - Cite the MUSDB18 Dataset</p> <pre><code>@misc{MUSDB18, author = {Rafii, Zafar and Liutkus, Antoine and Fabian-Robert St{\"o}ter and Mimilakis, Stylianos Ioannis and Bittner, Rachel}, title = {The {MUSDB18} corpus for music separation}, month = dec, year = 2017, doi = {10.5281/zenodo.1117372}, url = {https://doi.org/10.5281/zenodo.1117372} } </code></pre> <p>If compare your results with SiSEC 2018 Participants - Cite the SiSEC 2018 LVA/ICA Paper</p> <pre><code>@inproceedings{SiSEC18, author="St{\"o}ter, Fabian-Robert and Liutkus, Antoine and Ito, Nobutaka", title="The 2018 Signal Separation Evaluation Campaign", booktitle="Latent Variable Analysis and Signal Separation: 14th International Conference, LVA/ICA 2018, Surrey, UK", year="2018", pages="293--305" }</code></pre> <p>&nbsp;</p> <p>&nbsp;</p>