Table 1 Layerwise parameter of the deep convolution module.
From: Multiscale superpixel depth feature extraction for hyperspectral image classification
Layer type | Parameter |
|---|---|
The first Conv2D | \(128 \times 3 \times 3\), stride = 1 |
MaxPooling2D | \(2 \times 2\), stride = 2 |
Activation | ReLU |
The second Conv2D | \(64 \times 3 \times 3\), stride = 1 |
MaxPooling2D | \(2 \times 2\), stride = 2 |
Activation | ReLU |
