'same' padding with stride>1

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

The purpose of ‘same’ padding is to keep the dimensions of the image. Then what is meant by ‘same’ padding when the stride is greater than one?

2

Hi Meir,

As I understand it, this is done to make output and input size consistent when a particular stride is used. See, e.g.,

1 Like