Skip to content

Latest commit

 

History

History
28 lines (25 loc) · 3.35 KB

README.md

File metadata and controls

28 lines (25 loc) · 3.35 KB
Raw

DataSet-for-Vocal-Detection

The 2s music dataset is related of the FMA: A Dataset For Music Analysis, a dump of the Free Music Archive (FMA).

Data

  1. Jamendo Train : 26,715 tracks.
  2. Jamendo Test : 5,972 tracks.
  3. FMA-C-1 Train : 12,254 tracks.
  4. FMA-C-1 Test : 4,085 tracks.
  5. FMA-C-2 Train : 13,752 tracks.
  6. FMA-C-2 Test : 4,583 tracks.
  7. Test Hard : 8,291 tracks.
  8. A-Cappella : 7,922 tracks.
  9. Instrumental : 6,372 tracks.
  10. KTV Train : 6,534 tracks.
  11. KTV Test : 1,367 tracks.
  12. MIR-1k : 9,900 tracks.
  13. Chinese-CD Train : 5,223 tracks.
  14. Chinese-CD Test : 2,227 tracks.
  15. Taiwanese-CD Train : 1,240 tracks.
  16. Taiwanese-CD Test : 527 tracks.
  17. Taiwanese-stream Train : 4,149 tracks.
  18. Taiwanese-stream Test : 1,774 tracks.
  19. Classical Train : 5,733 tracks.
  20. Classical Test : 2,444 tracks.
  21. RapData : 391 tracks.