Music21 is an untrimmed video dataset crawled by keyword query from Youtube. It contains music performances belonging to 21 categories. This dataset is relatively clean and collected for the purpose of training and evaluating visual sound source separation models.
Source: Music Gesture for Visual Sound Separation
Image Source: https://towardsdatascience.com/midi-music-data-extraction-using-music21-and-word2vec-on-kaggle-cb383261cd4e
Variants: Music21
This dataset is used in 1 benchmark:
| Task | Model | Paper | Date |
|---|---|---|---|
| 10-shot image generation | song | 0-1 laws for pattern occurrences … | 2024-02-07 |
Recent papers with results on this dataset: