The DALI dataset

The DALI dataset is a large Dataset of synchronised audio, lyrics and notes for the audio full-duration, with – its time-aligned lyrics and – its time-aligned notes (of the vocal melody). Lyrics are described according to four levels of granularity: notes (and textual information un- derlying a given note), words, lines and paragraphs. For each song, we also provide additional multimodal information such as genre, language, musician, album covers or links to video clips. 

Go to https://github.com/gabolsgabs/DALI where you can find all the tools to work with the DALI dataset and a detailed description of how to use it.

For this version cite the article:

@article{meseguer2020creating,
title={Creating DALI, a Large Dataset of Synchronized Audio, Lyrics, and Notes},
author={Meseguer-Brocal, Gabriel and Cohen-Hadria, Alice and Peeters, Geoffroy},
journal={Transactions of the International Society for Music Information Retrieval}, volume={3}, number={1}, year={2020},
publisher={Ubiquity Press}
}

and the original paper:

@inproceedings{meseguer2019dali,
title={Dali: A large dataset of synchronized audio, lyrics and notes, automatically created using teacher-student machine learning paradigm},
author={Meseguer-Brocal, Gabriel and Cohen-Hadria, Alice and Peeters, Geoffroy},
journal={arXiv preprint arXiv:1906.10606},
year={2019}
}

This research has received funding from the French National Research Agency under the contract ANR-16-CE23-0017-01 (WASABI project)