Table 1.
Proposed convolutional neural network (CNN) model architecture.
| Layer (Type) | Output Shape | Parameters |
|---|---|---|
| input_1 (Input Layer) | (None, 180, 180, 3) | - |
| conv_1 (Conv2D) | (None, 180, 180, 64) | 1792 |
| batch_norm_1 (Batch Normalization) | (None, 180, 180, 64) | 256 |
| max_pool_1 (MaxPooling2D) | (None, 90, 90, 64) | 0 |
| conv_2 (Conv2D) | (None, 90, 90, 128) | 73,856 |
| conv_3 (Conv2D) | (None, 90, 90, 128) | 1,47,584 |
| batch_norm_2 (Batch Normalization) | (None, 90, 90, 128) | 512 |
| max_pool_2 (MaxPooling2D) | (None, 45, 45, 128) | 0 |
| conv_4 (Conv2D) | (None, 45, 45, 256) | 2,95,168 |
| conv5 (Conv2D) | (None, 45, 45, 256) | 5,90,080 |
| batch_norm_3 (Batch Normalization) | (None, 45, 45, 256) | 1024 |
| max_pool_3 (MaxPooling2D) | (None, 22, 22, 256) | 0 |
| conv_6 (Conv2D) | (None, 22, 22, 384) | 8,85,120 |
| conv_7 (Conv2D) | (None, 22, 22, 384) | 13,27,488 |
| batch_norm_4 (Batch Normalization) | (None, 22, 22, 384) | 1536 |
| max_pool_4 (MaxPooling2D) | (None, 11, 11, 384) | 0 |
| conv_8 (Conv2D) | (None, 11, 11, 480) | 16,59,360 |
| conv_9 (Conv2D) | (None, 11, 11, 480) | 20,74,080 |
| batch_norm_5 (Batch Normalization) | (None, 11, 11, 480) | 1920 |
| max_pool_5 (MaxPooling2D) | (None, 5, 5, 480) | 0 |