I have a question regarding Transpose Convolutions. What is the meaning of ‘same’ padding in these operations? In TF documentation it says that they are used to make sure that the output has the same height/width dimension as the input. However, when using transpose convolutions we are intentionally increasing these values from one layer to the next one, so I don’t understand what is the criteria for ‘same’ padding here.
Thanks in advance for your time and consideration.