Table 2.
CNN architecture 1 | CNN architecture 2 | ||
---|---|---|---|
Layers | Parameters | Layers | Parameters |
Input | 100×100 | Input | 100×100 |
Conv1 | 64×5×5,pad 0,stride 1 | Conv1 | 64×5×5,pad 0,stride 1 |
Leaky ReLU | alpha=0.01 | Leaky ReLU | alpha=0.01 |
Max Pool 1 | 3×3, stride 3, pad 0 | Max Pool 1 | 3×3, stride 3, pad 0 |
Conv2 | 64×2×2,pad 0,stride 1 | Conv2 | 64×2×2,pad 0,stride 1 |
Leaky ReLU | alpha=0.01 | Leaky ReLU | alpha=0.01 |
Max Pool 2 | 3×3, stride 3, pad 0 | Max Pool 2 | 3×3, stride 3, pad 0 |
Dropout | 0.1 | Dropout | 0.1 |
FC 1+ReLU | 128 | FC 1+ReLU | 128 |
FC 2+ReLU | 8 | LSTM 1 + ReLU | 8 |
L2 regularizer | 0.01 | L2 regularizer | 0.01 |
Dropout | 0.25 | Dropout | 0.25 |
FC 3+Sigmoid | 1 | FC 3 + Sigmoid | 1 |
Total parameters | 841,681 | Total parameters | 845,033 |