Table 3 LeNet-5 feature-extraction network.
Layer Name | Input size | Output size | Step size | Convolution kernel size | Pooled area |
|---|---|---|---|---|---|
Input layer | 32*32 | 28*28 | 1 | 5*5 | – |
ConV | 6@28*28 | 6@14*14 | 2 | – | 2*2 |
Pool | 6@14*14 | 16@10*10 | 1 | 5*5 | – |
ConV | 16@10*10 | 16@5*5 | 2 | – | 2*2 |
Pool | 16@5*5 | 120@1*1 | 1 | 5*5 | – |
FC | 1*120 | 1*84 | – | – | – |
FC | 1*84 | 1*7 | – | – | – |
Output | 1*7 | – | – | – | – |
