Table 9.
Comparative analysis of the proposed study with prior works.
| Study | Techniques | Accuracy (%) |
|---|---|---|
| [30] | ResNet50 feature extractor with SVM | 95.33 |
| [31] | SMOTE and ResNet152 with XGBoost and random forest | 97.70 |
| [32] | Customized CNN-based network | 84.22 |
| [33] | VGG-16-based scheme | 97.0 |
| [34] | Customized Xception Net | 95.0 |
| [35] | CNN with transfer multireceptive feature optimizer | 95.1 |
| [36] | Cascaded ResNet50V2 and Xception Net | 91.4 |
| [37] | Customized CNN-based model | 93.30 |
| [38] | Pre-trained deep learning models with GAN | 85.2 |
| Proposed | SWT + (AlexNet, ResNet101, and SqueezeNet) + iChi2 + SVM | 99.24 |
SVM: support vector machine; CNN: convolutional neural network; GAN generative adversarial network; SWT: stationary wavelet transform; iChi2: iterative chi-square.