Pitch-controlled Audio Generation on Song Describer Dataset (test)
163.24FDopenl3End-to-end
Evaluation Results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| End-to-endControl(s)=Pitch2026.03 | 163.24 | 0.75 | 32.75 | 0.03 | 173.1 | 35.61 | 3.7 | 4.5 | 3.1 | |
| ReadoutControl(s)=Pitch2026.03 | 107.45 | 0.54 | 37.08 | 0.038 | 15.7 | 5.66 | 3.5 | 3.5 | 1.3 | |
| LatCH-BControl(s)=Pitch2026.03 | 106.96 | 0.51 | 36.05 | 0.028 | 17.7 | 5.65 | 3.3 | 3.9 | 4.3 | |
| LatCH-FControl(s)=Pitch2026.03 | 105.57 | 0.59 | 35.82 | 0.041 | 17.7 | 5.65 | 3.6 | 3.7 | 3.3 | |
| SAOControl(s)=–2026.03 | 96.51 | 0.55 | 0.41 | 0.07 | 11.3 | 5.51 | — | — | — |