Single-layer convolutional network

1. How to construct the convolutional layer of the neural network. In the previous section we processed a 3D image by convolving two filters


In the above: There are 280 parameters. No matter how big the original image is, there are 280 parameters. The number of parameters is fixed.


Above: The number of channels in the output layer The number of filters used in this layer in the neural network. Then the number of channels in the filter must match the number of channels in the input.


Example:


A typical convolutional network has three layers: as follows:


Guess you like

Origin http://43.154.161.224:23101/article/api/json?id=325829553&siteId=291194637