Visualizing and understanding convnet---CNN visual comprehension

1. Definition of the wild ( receptive Field)

The definition and understanding of thereceptive fieldis thrown out here:

The feeling of the wild is actually a convolution neural network the size of the pixels on the original image of the pixel points on the feature map (feature map) of each layer output. Or the area of the input image that outputs the response of a node of a feature map is the field of perception.
For example, if our first layer is a 33 convolution core, then each node in the feature map we get through this convolution is derived from this 33 convolution core and the 33 area of the original image, then we call this Feature map 's node-sensing field size is 33
If you go through the pooling layer again, assuming that the stride of the convolution layer is 1,pooling layer size 22,stride is 2, then the pooling Layer node of the sensing field is 44
There are a few points to note that padding does not affect the feeling of the wild, stride only affect the next layer of featuremap feeling field, size affects the feeling of the layer of the wild.
The other definition of this is: in convolutional neural network CNN, the area of the input layer corresponding to an element in a certain layer of output is called the receptive field.
???? Here is a partial link for reference:
???? 1. The Field of sensation in deep neural networks (receptive fields)
???? 2. What's a receptive field in a convolutional neural network?
???? 3. Painless understanding of the receptive field in CNN
???? 4. Visualizing what Convnets Learn---cs231n
???? 5. Deep visualization: Visualize and understand CNN
???? 6. CNN receptive Field Calculator--CNN Feel Wild Calculator
? ? ? ?

Visualizing and understanding convnet---CNN visual comprehension