Table 9 Performances of TOPCONet model and its sub-modules along with some state-of-the-art models on COVIDx Dataset.
From: TOPSIS aided ensemble of CNN models for screening COVID-19 in chest X-ray images
Work Ref. | Technique | #Trainable parameters | Performance (in %) in terms of | |||
|---|---|---|---|---|---|---|
Precision | Recall | F1-score | Accuracy | |||
Jain et al.57 | Xception | 22,910,480 | 79.50 | 74.00 | 73.00 | 74.25 |
Chowdhury et al.56 | CheXNet | 20,242,984 | 82.50 | 75.00 | 73.50 | 74.75 |
Bashar et al.34 | Optimized CNN | 138,357,544 | 87.50 | 85.50 | 85.00 | 85.50 |
Senan et al.35 | ResNet50 | 25,636,712 | 80.50 | 77.00 | 76.50 | 77.25 |
Dey et al.19 | CovidConvLSTM | 363,996,809 | 88.71 | 86.75 | 86.58 | 86.75 |
Proposed | Classifier 1 | 306,467 | 94.00 | 92.61 | 93.30 | 93.25 |
Proposed | Classifier 2 | 339,235 | 92.00 | 92.46 | 92.23 | 92.25 |
Proposed | Classifier 3 | 355,619 | 81.00 | 69.53 | 74.83 | 72.75 |
Proposed | TOPCONet | 1,001,324 | 94.50 | 94.03 | 94.26 | 94.25 |
