Table 4 Classification comparison with different attention models.
From: Brain CT image classification based on mask RCNN and attention mechanism
Model | Accuracy/% | Precision/% | Recall/% | F1/% | Kappa/% |
|---|---|---|---|---|---|
CBAM-DResNet50 | 89.38 | 85.26 | 81.76 | 83.41 | 80.83 |
TCCBAM-DResNet50 | 89.73 | 86.78 | 87.82 | 87.22 | 82.19 |
CBAM-DResNet34 | 89.96 | 87.25 | 88.93 | 88.05 | 82.45 |
TCCBAM-DResNet34 | 90.51 | 88.47 | 89.61 | 88.97 | 83.62 |
CBAM-DResNet18 | 90.82 | 89.67 | 90.57 | 80.11 | 84.75 |
TCCBAM-DResNet18 | 91.55 | 90.63 | 91.49 | 91.18 | 84.98 |
CBAM-DResNet10 | 95.37 | 93.69 | 94.18 | 93.85 | 85.31 |
TCCBAM-DResNet10 | 98.06 | 96.85 | 97.64 | 97.38 | 88.79 |
