2D Global max pooling
The 2D Global max pooling block performs exactly the same operation as the 2D Max pooling block except that the pool size (i.e., Horizontal pooling factor x Vertical pooling factor) is the size of the entire input of the block, i.e., it computes a single max value for all the incoming data.
The 2D Global max pooling block takes a tensor of size (input width) x (input height) x (input channels) and computes the max value of all values across the entire (input width) x (input height) matrix for each of the (input channels). The output is thus a tensor of size is 1 x 1 x (input channels).
Use 2D Global max pooling blocks as an alternative to the Flattening block after the last pooling block of your convolutional neural network. Using 2D Global max pooling block can replace the fully connected blocks of your CNN. For more information, see Section 3.2 of Min Lin, Qiang Chen, Shuicheng Yan. Network In Network.