0
$\begingroup$

Maxpooling is performed as one of the steps in inception which yields same output dimension as that of the input. Can anyone explain how this max pooling is performed?

$\endgroup$
1
$\begingroup$

My bad, didn't look into the block diagram first

enter image description here

maxpooling filter/kernel is 3x3 = fxf

Formula for padding to get the same dimension as the input

p = (f - 1)/2

here f=3 ,so padding to be performed before maxpooling is (3-1)/2 = 1

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.