Audio Classification on FSD50K
Metric: mAP (higher is better)
Metric: mAP (higher is better)
| # | Model↕ | mAP▼ | Extra Data | Paper | Date↕ | Code |
|---|---|---|---|---|---|---|
| 1 | ONE-PEACE | 69.7 | Yes | ONE-PEACE: Exploring One General Representation ... | 2023-05-18 | Code |
| 2 | MN | 65.6 | Yes | Dynamic Convolutional Neural Networks as Efficie... | 2023-10-24 | Code |
| 3 | PaSST-S | 65.55 | Yes | Efficient Training of Audio Transformers with Pa... | 2021-10-11 | Code |
| 4 | DyMN-L | 65.5 | Yes | Dynamic Convolutional Neural Networks as Efficie... | 2023-10-24 | Code |
| 5 | PaSST-N-S | 64.2 | Yes | Efficient Training of Audio Transformers with Pa... | 2021-10-11 | Code |
| 6 | PSLA | 56.71 | Yes | PSLA: Improving Audio Tagging with Pretraining, ... | 2021-02-02 | Code |
| 7 | MATPAC (SSL Model) | 55.2 | No | Masked Latent Prediction and Classification for ... | 2025-02-17 | Code |
| 8 | Temporal Knowledge Distillation for On-device Audio Classification | 54.8 | No | Temporal Knowledge Distillation for On-device Au... | 2021-10-27 | - |
| 9 | Large 6-Layer Transformer with Pooling | 53.7 | No | Audio Transformers | 2021-05-01 | - |
| 10 | [ABT] AudioNTT | 0.474 | No | Audio Barlow Twins: Self-Supervised Audio Repres... | 2022-09-28 | Code |