search

forresti / SqueezeNet

SqueezeNet: AlexNet-level accuracy with 50x fewer parameters
BSD 2-Clause "Simplified" License
2.17k stars 723 forks source link

v1.1 dimensions mismatch in the first layers #33

Closed gudovskiy closed 7 years ago

gudovskiy commented 7 years ago

Hi, seems that original v1.0 had nice dimension relationships: 227 -> (227-7)/2+1=111 -> (111-3)/2+1=55 etc. But in v1.1 we start to get: 227 -> (227-3)/2+1=113 -> (113-3)/2+1=56 etc.

To get output of conv1 to be 111, input image should be decreased to 223x223. Not sure how exactly Caffe handles this, but something mismatch in v1.1. Any idea?

forresti commented 7 years ago

You can look in the conv_layer.cpp code in Caffe for the details of the rounding scheme. But, yes, there is a rounding scheme, and it all works nicely.