Research Article
[Retracted] Detecting Digital Watermarking Image Attacks Using a Convolution Neural Network Approach
Table 1
Important parameters of the CNN model.
| Parameters | Values |
| Kernel size | 3 | Max-pooling size | 2 | Drop out | 0.50 | Fully connected layer | 128 | Activation function | ReLu function | Optimizer | Adam | Epochs number | 20 | Batch size number | 50 |
|
|