Automatic Chord Estimation on ACR (test)
83.03Root AccuracyBTC (CL)
Evaluation results
| Method | Links | |||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| BTC (CL)Labels=Full labels, Strategy=Data-Incremental Continual Learning, KD alpha=0.32026.02 | 83.03 | 80.17 | 78.33 | 69.38 | 67 | 80.24 | 80.16 | 80.71 | 87.1 | 65.6 | 51.6 | 42.3 | 55.7 | 23.5 | 45.6 | 18.5 | 25.5 | 71.6 | 49.5 | |
| BTC (CL)Labels=50%, Strategy=Data-Incremental Continual Learning, KD alpha=0.32026.02 | 82.14 | 78.58 | 76.67 | 66.77 | 64.28 | 78.62 | 78.87 | 80.17 | 84.5 | 61.2 | 50.2 | 42.1 | 51.6 | 18.3 | 47.4 | 14.8 | 22.3 | 68.7 | 47.1 | |
| 2E1D (CL)Labels=Full labels, Strategy=Data-Incremental Continual Learning, KD alpha=0.32026.02 | 81.55 | 78.52 | 76.85 | 68.19 | 65.49 | 78.98 | 78.52 | 80.73 | 90 | 68.9 | 39.4 | 35.3 | 51.9 | 14.8 | 46.4 | 8.4 | 18.8 | 71.9 | 33.7 | |
| BTC (SL)Parameters=3.03M, Strategy=Traditional supervised learning, Labels=Full labels2026.02 | 81.52 | 78 | 76.12 | 65.93 | 63.44 | 78.11 | 77.79 | 78.38 | 88.7 | 63.1 | 36.7 | 28.2 | 52.8 | 23.3 | 0 | 5.9 | 22.9 | 68.6 | 39 | |
| 2E1D (CL)Labels=50%, Strategy=Data-Incremental Continual Learning, KD alpha=0.32026.02 | 81.02 | 77.54 | 75.76 | 66.39 | 63.68 | 77.93 | 77.57 | 80.23 | 90.9 | 63 | 33.3 | 29.5 | 49 | 12.7 | 46.7 | 5.2 | 16.6 | 70.4 | 31.2 | |
| 2E1D (SL)Parameters=2.2M, Strategy=Traditional supervised learning, Labels=Full labels2026.02 | 79.39 | 76.14 | 74.53 | 64.53 | 61.67 | 76.89 | 76.24 | 78.5 | 91.7 | 62.1 | 23.9 | 25.2 | 44.4 | 3.5 | 0 | 0 | 8 | 68.7 | 24.4 | |
| Li et al.Framework=Noisy Student2026.02 | 78.05 | 73.56 | 70.71 | 61.11 | 56.84 | 74.64 | 74.26 | — | — | — | — | — | — | — | — | — | — | — | — | |
| Bortolozzo et al.Framework=Noisy Student2026.02 | — | — | — | — | — | — | — | — | 55.5 | 54.4 | 51 | 12 | 47.5 | 30.4 | 43.3 | — | — | 46.8 | 24.5 |