Table 1.
Comparison of the proposed HOG + CNN model with related works using chest X-ray images
| Author | Architecture | CXR Images/class | Accuracy % |
|---|---|---|---|
| Razzak et al. [20] | MobileNet |
COVID-19 (200) Viral pneumonia (200) Bacterial pneumonia (200) Normal (200) |
80.95 |
| Asif and Wenhui [21] | Inception V3 |
COVID-19 (864) Pneumonia (1345) Normal (1341) |
96 |
| Pathari and Rahul [22] | MobileNet V3 |
COVID-19 (6000) Pneumonia (6000) Normal (6000) |
95.58 |
| Makris et al. [23] | VGG16 |
COVID-19 (112) Pneumonia (112) Normal (112) |
95.88 |
| Gomes et al. [24] | SVM |
COVID-19 (464) Viral pneumonia (1490) Bacterial pneumonia (2783) Normal (1583) |
89.78 |
| Elaziz et al. [26] | K Nearest Neighbor (KNN) |
COVID-19 (216) Normal (1675) |
96.09 |
| Wang and A. Wong [27] | Deep Convolutional Neural Network (COVID-Net) |
COVID-19 (358) Pneumonia (5538) Normal (8066) |
93.3 |
| Duran-Lopez et al. [28] | Deep Convolutional Neural Network (COVID-XNet) |
COVID-19 (2589) Normal (4337) |
94.43 |
| Oh et al. [29] | ResNet-18 |
Viral Pneumonia + COVID-19 (200) Bacterial pneumonia (54) Tuberculosis (57) Normal (191) |
88.9 |
| Proposed HOG + CNN Model | Deep Convolutional Neural Network (HOG + CNN) |
COVID-19 (576) Pneumonia (4273) Normal (1583) |
96.74 |