Table 4 Comparison of the proposed model with the literature.
Author | Year | Technique | Use of XAI | Accuracy (%) | Misclassification rate (%) |
|---|---|---|---|---|---|
Ali et al.47 | 2022 | VGG16 ResNet50 InceptionV3 Ensemble | No | 81.48 ± 6.87 82.96 ± 4.57 74.07 ± 3.78 79.26 ± 1.05 | 18.52 ± 6.87 17.04 ± 4.57 25.93 ± 3.78 20.74 ± 1.05 |
Burak Gülmez48 | 2022 | Ensemble approach | No | 84.2 | 15.8 |
Irmak et al.49 | 2022 | MobileNetV2 VGG16 VGG19 | No | 91.38 83.62 78.45 | 8.62 16.38 21.55 |
Singh and Songare 50 | 2022 | VGG-16 ResNet50 InceptionV3 GoogLeNet | No | 83.85 85.38 86.37 88.27 | 16.15 14.62 13.63 11.73 |
Sharma et al.51 | 2023 | ResNet18-based model | No | 84.59 | 15.41 |
Sethy et al.52 | 2023 | Darknet 19 Improved Darknet 19 | No | 81.4 85.49 | 18.6 14.51 |
Uysal53 | 2023 | CNN-LSTM hybrid model | No | 87 | 13 |
Ariansyah et al.54 | 2023 | CNN VGG16 | No | 64.52 83.33 | 35.48 16.67 |
Kundu et al.55 | 2023 | SVM KNN RestNet50 ViT | No | 65 84 91 93 | 35 16 9 7 |
Aqsa Akram et al.56 | 2024 | SkinMarkNet | No | 90.615 | 9.385 |
Proposed model | 2024 | VGG16 empowered with LRP | Yes | 93.29 | 6.71 |
