Table 2. Comparing performance between the baselines and the proposed method based on various evaluation metrics and corresponding standard deviations.
F32_R50 and F32_M denotes FCN32_ResNet50 and FCN32_MobileNet while EMV, EECE, EMCE), and EMVEM denotes Ensemble (Majority Voting (MV), EMV), Ensemble (Weighted Voting (ECE), EECE), Ensemble (Weighted Voting (MCE), EMCE), and Ensemble (Majority Voting + ECE + MCE (MVEM), EMVEM).
| DL | Accuracy (%) | Sensitivity (%) | Specificity (%) | F1score (%) | ECE (%) | MCE (%) |
|---|---|---|---|---|---|---|
| UNet | 95.4±2.5 | 90.7±3.9 | 88.9±4.5 | 93.4±3.0 | 3.2±1.3 | 39.7±18.8 |
| PSPNet | 95.0±2.0 | 89.1±3.9 | 88.2±4.3 | 92.5±2.9 | 4.6±1.2 | 40.6±14.9 |
| FCN32 | 95.8±2.4 | 92.3±4.5 | 91.0±5.0 | 94.0±3.5 | 2.5±2.1 | 37.6±19.1 |
| F32_R50 | 96.0±2.5 | 92.3±5.5 | 91.4±5.9 | 94.3±3.8 | 2.3±2.3 | 29.8±20.3 |
| F32_M | 95.2±2.3 | 91.0±4.7 | 90.1±5.6 | 93.1±3.3 | 4.1±1.6 | 38.2±19.9 |
| EMV | 98.8±0.6 | 94.1±3.0 | 92.9±3.6 | 96.6±1.7 | 2.4±1.2 | 28.1±14.1 |
| EECE | 99.1±0.5 | 95.4±2.9 | 94.3±2.9 | 97.1±1.5 | 2.3±1.2 | 24.7±12.4 |
| EMCE | 98.7±0.7 | 93.9±3.7 | 92.6±3.7 | 96.3±1.9 | 2.4±1.2 | 28.9±14.6 |
| EMVEM | 99.2±0.4 | 97.7±2.3 | 95.4±2.3 | 98.4±0.8 | 2.1±1.1 | 20.1±10.1 |