FSD50K
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Environmental sound classification | FSD50K | mAP69.7 | 126 | |
| Sound Event Tagging | FSD50K (test) | mAP69.7 | 42 | |
| Audio Tagging | FSD50K (eval) | mAP94 | 19 | |
| Speech Enhancement | FSD50K Clarity | PESQ4.64 | 14 | |
| Audio Classification | FSD50K (train/test) | mAP58.27 | 9 | |
| Text-to-Audio Generation | FSD50K | FD25.97 | 8 | |
| Acoustic Scene Classification | FSD50K (test) | mAP54.1 | 8 | |
| Hallucination Detection | FSD50K | FPR1 | 4 | |
| Audio Classification | FSD50K NAS-Bench-360 | Runtime (GPU hours)4 | 4 | |
| Audio-to-Text Retrieval | FSD50K | R@169.7 | 3 | |
| Ambisonics Encoding | FSD50K unseen (test) | SI-SDR-7.31 | 2 |
Showing 11 of 11 rows