deezer/MultilingualLyricsToAudioAlignment

DALI datasets split used to train models presented in the paper Multilingual lyrics-to-audio alignment (ISMIR 2020).

MultilingualLyricsToAudioAlignment

DALI dataset split, i.e. sng_id for each dataset, (https://github.com/gabolsgabs/DALI) used to train models presented in the paper Multilingual lyrics-to-audio alignment (ISMIR 2020).