Table 18 Comparative analysis of recent deep learning models vs. Our proposed Model.
From: Attention-Enhanced CNNs and transformers for accurate monkeypox and skin disease detection
Ref. No. | Model Type | Dataset | Accuracy (%) | Precision (%) | Recall (%) | F1-Score (%) | AUC (%) |
|---|---|---|---|---|---|---|---|
Ensemble CNN (Inception V3, Xception, DenseNet169) | Public monkeypox dataset | 93.39 | 88.91 | 96.78 | 92.35 | 95.5 | |
CNN optimized with Grey Wolf Optimizer (GWO) | Augmented monkeypox dataset | 95.3 | - | - | - | 96.0 | |
Hyper-parameter tuned Yolov5 | Roboflow dataset | 98.18 | - | - | - | 98.5 | |
Secure CNN (DarkNet-53 with cancelable biometrics) | DarkNet-53 trained dataset | 98.81 | 98.9 | 97.02 | 97.95 | 99.0 | |
Automated deep feature engineering model | 910 open-source images | 91.87 | - | - | - | - | |
Deep learning for multi-class skin disease classification | Public dataset | High accuracy | - | - | - | - | |
AI-driven real-time monitoring system | Public dataset | 94.51 | 99.3 | 94.1 | 96.6 | 99.5 | |
Computational model for outbreak prediction | Epidemiological data | 97.25 | - | - | - | - | |
(Our Work) | EfficientNetB7 + Coordinate Attention | Enhanced monkeypox dataset | 99.99 | 99.8 | 99.9 | 99.85 | ~ 100 |
