Table 1 Classic 2D-CNN layers architecture
Layers | Types | Output | Parameters |
|---|---|---|---|
Input | Input Layer | (None, 64, 64, 3) | 0 |
Block_1 | Conv2D_1 | (None, 64, 64, 16) | 1216 |
BatchNormalization_1 | (None, 64, 64, 16) | 64 | |
Activation_1 (ReLU) | (None, 64, 64, 16) | 0 | |
MaxPooling2D_1 | (None, 32, 32, 16) | 0 | |
Dropout_1 | (None, 32, 32, 16) | 0 | |
Block_2 | Conv2D_2 | (None, 32, 32, 32) | 4640 |
BatchNormalization_2 | (None, 32, 32, 32) | 128 | |
Activation_2 (ReLU) | (None, 32, 32, 32) | 0 | |
MaxPooling2D_2 | (None, 16, 16, 32) | 0 | |
Dropout_2 | (None, 16, 16, 32) | 0 | |
Block_3 | Conv2D_3 | (None, 16, 16, 64) | 8256 |
Conv2D_4 | (None, 16, 16, 64) | 16,448 | |
BatchNormalization_3 | (None, 16, 16, 64) | 256 | |
Activation_3 (ReLU) | (None, 16, 16, 64) | 0 | |
MaxPooling2D_3 | (None, 8, 8, 64) | 0 | |
Dropout_3 | (None, 8, 8, 64) | 0 | |
Head | Flatten | (None, 4096) | 0 |
Dense_1 | (None, 64) | 262,208 | |
Dropout_4 | (None, 64) | 0 | |
Output | Dense_2 | (None, 1) | 65 |
Sigmoid | (None, 1) | 0 |