Table 1.
Summarization of related works for stand-alone DL models.
Table 2.
Summarization of related works for hybrid DL models.
Fig 1.
Radiography X-ray image from the dataset.
Table 3.
Summary of our experimental dataset splitting into training and testing images.
Table 4.
Deep residual learning for image recognition [34].
Fig 2.
Proposed RVCNet architecture.
Fig 3.
Basic methodology of the overall system.
Table 5.
Performance comparison with different lea rning rates, batch sizes and optimizer functions.
Table 6.
10-fold cross-validation metrics of individually executed iterations.
Table 7.
Results of multiple runs for RVCNet.
Table 8.
Summary table with the mean, standard variation, best, worst, count, and total for the given metrics across all 5 runs.
Table 9.
Comparison of proposed architecture with separate models.
Fig 4.
Training and Testing analysis of RVCNet within 25 epochs: (a) for accuracy (b) for loss.
Fig 5.
Confusion matrix of proposed RVCNet.
Fig 6.
The ROC curves of the proposed RVCNet.
Fig 7.
Visualization for COVID-19 X-ray test images with Grad-CAM heatmaps: (a, c) original images; (b, d) corresponding Grad-CAM heatmaps.
Fig 8.
Visualization for Lung Opacity X-ray test images with Grad-CAM heatmaps: (a, c) original images of four samples; (b, d) corresponding Grad-CAM heatmaps.
Fig 9.
Visualization for Viral Pneumonia X-ray test images with Grad-CAM heatmaps: (a, c) original images; (b, d) corresponding Grad-CAM heatmaps.
Table 10.
Comparison of RVCNet with others for several data samples.
Table 11.
Comparison of RVCNet with others for classification of COVID, viral pneumonia, lung opacity, and healthy persons.