Research Article
Convolutional Neural Network-Based Fish Posture Classification
Table 1
The parameters of convolutional neural networks.
| Layer | Conv1 | Conv2 | Conv3 | Conv4 | Conv5 | Conv6 | FC1 | FC2 |
| Input size | | | | | | | | 512 | Convolution kernel | | | | | | | | | Output | 32 | 64 | 64 | 128 | 128 | 128 | | | Boundary filling | 1 | 1 | 1 | 1 | 1 | 1 | | | Step | 1 | 1 | 1 | 1 | 1 | 1 | | | Activation function | ReLU | ReLU | ReLU | ReLU | ReLU | ReLU | ReLU | ReLU | Pooling | | | | | | | | | Dropout | | | | | | | 0.6 | |
|
|