Average Pooling returns the average of all the values from the portion of the image covered by the Kernel.

Average Pooling in Convolutional Deep Learning

Max Pooling returns the maximum value from the portion of the image covered by the Kernel.

Max Pooling in Convolutional Deep Learning

University of Michigan - Ann Arbor

Pooling is the process of down-sampling, i.e., reducing the size of the matrix. A filter is passed over the results of the previous layer and outputs one number out of each group of values.

Pooling Layer in Convolutional Deep Learning

When input size is $n_H * n_w * n_c$. the padding output size would be $floor((n_H - f)/s +1) * floor((n_W - f)/s + 1) * n_C$

$n_H$, $n_W$  and  $n_C$  are the hight , wight and channels of the input layer.



The output size of pooling

https://towardsdatascience.com/classifying-skin-lesions-with-convolutional-neural-networks-fc1302c60d54

Classifying Skin Lesions with Convolutional Neural Networks

- It decreases the computational power required to process the data through dimensionality reduction.
- It is useful for extracting dominant features which are rotational and positional invariant, thus maintaining the process of effective model training.

Advantages of Pooling in Convolutional Deep Learning

Types of Pooling Layer in Convolutional Deep Learning

- Filter size
- Stride
- Max or Average pooling
- Typically, padding size = 0.

No parameters are needed to learn for pooling because of the fixed function (max or average).

Pooling Parameters and Hyperparameters in Convolutional Deep Learning

1. Reduce parameters. By reducing the dimension of feature map, the parameters required for subsequent layers are effectively reduced

2. Translation variance. It means that for input, when the pixels in the neighborhood are slightly displaced, the output of the pooling layer is unchanged. This enhances the robustness of the network and has a certain anti disturbance effect (translation invariance)

The error of feature extraction mainly comes from two aspects:

(1) The variance of the estimated value increases due to the limited size of the neighborhood;

(2) The convolution layer parameter error causes the offset of the estimated mean.

Generally speaking, mean pooling can reduce the first error and retain more background information of the image. Max pooling can reduce the second error and retain more texture information.

Learn Before

Related

Learn After