Why does the network in the network continuously connect two convolutions of 1* 1?
1* 1 convolution filter is the same as ordinary filter, the only difference is that its size is 1* 1, and the relationship between the local information of the previous layer is not considered. It first appeared in the paper Network in Network. The purpose of using 1* 1 convolution is to deepen and broaden the network structure, and use it to reduce the dimension when using convolution for deeper processing.