Multi-label Classification on MuReD (test)
57.3ML F1C-Tran
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| C-TranBackbone=DenseNet161, Optimizer=Adam, Learning Rate=10-5, Loss=Polynomial loss, Image Size=384x384, Batch Size=32, Algorithm=LP ROS (10% resampling ratio)2022.07 | 57.3 | 68.5 | 96.2 | 82.4 | 97.6 | 82.4 | 90 | |
| Wang et al.Ensemble=VGG16 and EfficientNetB3, Loss=WBCE, Pre-processing=CLAHE2022.07 | 31.5 | 37.9 | 84.5 | 61.2 | 89.7 | 66.9 | 75.4 | |
| RIADD 1stEnsemble=EfficientNetB5 and B62022.07 | 20.8 | 38 | 88.1 | 63 | 89.3 | 63.7 | 76.2 | |
| Gour et al.Ensemble=VGG16 and EfficientNetB32022.07 | 1 | 26.2 | 82.5 | 54.4 | 87.2 | 50.7 | 70.8 |