Audio Classification on AIR-Bench Speech
47.16Emotion Acc (MELD)MTC (AudioSet)
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MTC (AudioSet)Training Data=AudioSet2026.03 | 47.16 | 25.9 | 47.27 | 45.06 | 37.24 | |
| Contrastive-scratch (Ours)Training Data=Caption*2026.03 | 44.08 | 41.34 | 73.11 | 77.82 | 40.48 | |
| Multi-Task (Ours)Training Data=Caption*, Tag*2026.03 | 41.7 | 32.86 | 53.38 | 57 | 65.47 | |
| MTC (Ours-UTS)Training Data=UTS2026.03 | 39.19 | 48.79 | 74.45 | 92.64 | 86.59 | |
| Contrastive-scratch [78]Training Data=Original Captions2026.03 | 29.61 | 48.79 | 55.25 | 84.85 | 86.59 | |
| Captioning-scratch (Ours)Training Data=Caption*2026.03 | 23.47 | 29.9 | 46.4 | 28.02 | 18.82 | |
| Captioning-scratch [78]Training Data=Original Captions2026.03 | 22.93 | 37.4 | 70.2 | 78.93 | 40.14 | |
| PAR (Ours-UTS)Training Data=UTS2026.03 | 17.9 | 38.64 | 47.82 | 37.4 | 57.16 |