Table 1 Comparison of recent deep Learning-Based approaches for Monkeypox Detection.
From: Attention-Enhanced CNNs and transformers for accurate monkeypox and skin disease detection
Ref No. | Main Contribution | Methodology | Dataset | Performance Metrics | Key Findings |
|---|---|---|---|---|---|
Ensemble learning framework for monkeypox detection | Pre-trained CNNs (Inception V3, Xception, DenseNet169) with Beta function normalization | Public monkeypox skin lesion dataset (five-fold cross-validation) | Accuracy: 93.39%, Precision: 88.91%, Recall: 96.78%, F1-score: 92.35% | Ensemble learning improves classification accuracy significantly | |
Optimized CNN classification using GWO | CNNs optimized with Grey Wolf Optimizer | Monkeypox skin lesion dataset with additional augmentation | Accuracy: 95.3%, F1-score: Improved with GWO | CNNs optimized with GWO outperform standard CNNs | |
Hyper-parameter tuning in deep learning for monkeypox detection | Transfer learning and hyper-parameter tuning with Yolov5 | Roboflow skin lesion dataset | Accuracy: 98.18% (Yolov5) | Hyper-parameter tuning boosts accuracy in skin lesion classification | |
Secure monkeypox classification using CanDark model | Cancelable biometrics integrated with DarkNet-53 CNN | DarkNet-53 trained on monkeypox dataset | Accuracy: 98.81%, Specificity: 98.73%, Precision: 98.9%, Recall: 97.02% | Secure framework with high accuracy and biometric data protection | |
Automated monkeypox detection model using deep feature engineering | Deep feature engineering using multi-step classification pipeline | 910 open-source images (five disease categories) | Accuracy: 91.87% | Automated pipeline effectively classifies monkeypox and similar diseases | |
Deep learning-based classification of skin diseases including monkeypox | Deep learning-based multi-class classification of skin diseases | Public dataset of skin disease images including monkeypox | Accuracy: High classification accuracy for multi-class skin diseases | Deep learning proves effective for multi-class skin disease classification | |
Real-time AI-based monkeypox detection using ‘Super Monitoring’ | AI and cloud-based real-time monitoring system | Publicly accessible datasets for real-time monitoring | Accuracy: 94.51% across six skin disorders | AI-based real-time monitoring enhances early detection | |
Federated learning for secure classification of monkeypox | Federated learning applied to deep CNNs | Federated data from multiple sources | Federated learning achieved high classification performance | Federated learning ensures secure and efficient classification | |
Computational model for predicting monkeypox outbreaks | Epidemiological data analysis using computational models | Epidemiological records from WHO and CDC | Prediction accuracy: 97.25% for outbreak scenarios | Computational models help in predicting outbreaks with high accuracy |
