1. Layer 1 in a Convolutional Neural Network
Add different bias b to the results of each convolution kernel, and then perform nonlinear activation respectively
The last dimension of a is the same as the number of convolution kernels
2. No matter how big the input image is, the number of parameters of the convolution layer (related to the number of convolution kernels) is fixed
3. Symbol description
For the convolutional layer l
1) Indicates the number of channels of the output image of the previous layer, that is, the number of channels of the convolution kernel of this layer
2) Represents the size of the convolution kernel: f-[l] * f-[l] * n_C
3) Indicates the number of pixels filled in the one-sided edge of the image
4) Represents the stride of moving the convolution kernel on the image each time
5) The dimension of the input image: , where H stands for Height; W stands for Width
6) Dimensions of the output image:
,
7) with
8) Indicates the number of convolution kernels in this layer, and the size of each convolution kernel is
9) The dimension of the activation result:
For m samples:
10) The number of parameters w:
11) The number of offset b:
4. Sometimes the order of dimensions changes (doesn't affect anything)