Table 2.
Layer Type | Filter Size@ Kernel Size |
Activation Function | Output Shape | Number of Parameters |
---|---|---|---|---|
Input 1D | 1 × 300 | |||
Conv 1D_1 | 128 @ 1 × 3 | ReLu | 1 × 149 | 512 |
BatchNorm 1D_1 | 1 × 149 | 256 | ||
Max pool 1D_1 | 1 × 2 | 1 × 148 | ||
Conv 1D_2 | 128 @ 1 × 3 | ReLu | 1 × 73 | 49,280 |
BatchNorm 1D_2 | 1 × 73 | 256 | ||
Max pool 1D_2 | 1 × 2 | 1 × 72 | ||
Conv 1D_3 | 128 @ 1 × 3 | ReLu | 1 × 35 | |
BatchNorm 1D_3 | 1 × 35 | 49,280 | ||
Max pool 1D_3 | 1 × 2 | 1 × 34 | 256 | |
Input 2D | 123 × 300 | |||
Conv 2D_1 | 128 @ 3 × 3 | ReLu | 61 × 149 | |
BatchNorm 2D_1 | 61 × 149 | 1280 | ||
Max pool 2D_1 | 2 × 2 | 60 × 148 | 256 | |
Conv 2D_2 | 128 @ 3 × 3 | ReLu | 29 × 73 | |
BatchNorm 2D_2 | 29 × 73 | 147,584 | ||
Max pool 2D_2 | 2 × 2 | 28 × 72 | 256 | |
Conv 2D_3 | 128 @ 3 × 3 | ReLu | 13 × 35 | |
BatchNorm 2D_3 | 13 × 35 | 147,584 | ||
Max pool 2D_3 | 2 × 2 | 12 × 34 | 256 | |
1D FC_1 | ReLu | 4352 | 43,530 | |
2D FC_2 | ReLu | 52,224 | 522,250 | |
FC_3 | 10 | |||
Output | 2 | 42 |