Research Article
An Efficient Method for Diagnosing Brain Tumors Based on MRI Images Using Deep Convolutional Neural Networks
Table 2
Some CNN architectures and ImageNet benchmark (image classification).
| CNN architectures | Author | Top 1 accuracy (%) | Top 5 accuracy (%) |
| AlexNet | Krizhevsky et al. [26] | 63.3 | 84.6 | VGG16 | Simonyan and Zisserman [7] | 74.4 | 91.9 | VGG19 | 74.5 | 92.0 | GoogLeNet (InceptionV1) | Szegedy et al. [27] | 69.8 | 89.5 | ResNet50 | He et al. [5] | 77.1 | 93.3 | ResNet152 | 78.6 | 94.3 | DenseNet201 | Huang et al. [4] | 77.4 | 93.7 | MobileNet224 | Howard et al. [6] | 70.6 | 89.5 | MobileNetV3 | 79.0 | 94.5 | ResNeXt101 | Xie et al. [28] | 80.9 | 95.6 | EfficientNet-L2 | Tan and Le [29] | 90.2 | 98.8 | RegNet-Y | Radosavovic et al. [30] | 79.9 | 95.0 |
|
|