Kara1k dataset

dataset for Music Information Retrieval

Yann Bayle, Ladislav Maršík, Martin Rusek


Features from professional audio tracks (vocal / instrumental) of cover songs provided by Recisio's KaraFun application versus the original songs by the original authors. The dataset is dedicated for Cover Song Identifiation and Singing voice analysis.

Where to get it



Contact email:



Supporting research projects and grants:

GAUK 708314


  • Bayle Y., Maršík L., Rusek M., Robine M., Hanna P., Slaninová K., Martinovič J., Pokorný J.: Kara1k: a karaoke dataset for cover song identification and singing voice analysis, in 2017 IEEE International Symposium on Multimedia (ISM), Taichung, Taiwan, IEEE, ISBN: 978-1-5386-2937-6, pp. 177-184, 2017 - text
The content of this web site is licensed under Creative Commons Attribution-NonCommercial 3.0 Czech Republic