Acoustic Scene Classification on UAS DCASE Task 1A 2018 (test)
78.1AccuracyERGL
Evaluation Results
| Method | Links | |
|---|---|---|
| ERGLModel structure=CNN and Graph Learning2022.10 | 78.1 | |
| ERGLApproach=one-way event-to-scene inference2022.10 | 78.08 | |
| Relation-guided ASC (RGASC)Model structure=two-tower model2022.10 | 77.35 | |
| Wavelet-based spectrumModel structure=CRNN2022.10 | 76.6 | |
| MLTFModel structure=CNN and SVM2022.10 | 75.3 | |
| PANNModel structure=VGG-like CNN, Transfer learning mode=Fine-tuning mode2022.10 | 73.8 | |
| ABCNNModel structure=Attention-Based CNN2022.10 | 72.6 | |
| Model with AttentionModel structure=CRNN with Self-attention2022.10 | 70.8 | |
| NNF_CNNEnsModel structure=CNN and nearest neighbor filters2022.10 | 69.3 | |
| CNN_from_SurreyModel structure=CNN2022.10 | 68 | |
| Conditional scene and event recognitionLoss type=conditional loss2022.10 | 66.39 | |
| MTL-based event and scene analysisFramework=Multi-task learning2022.10 | 61.69 | |
| BaselineModel structure=CNN2022.10 | 59.7 | |
| PANNModel structure=VGG-like CNN, Transfer learning mode=Fixed mode2022.10 | 56.9 | |
| Scene and event jointly classification2022.10 | 52.35 |