search

wkentaro / pytorch-fcn

PyTorch Implementation of Fully Convolutional Networks. (Training code to reproduce the original result is available.)
MIT License
1.74k stars 480 forks source link

Why didn't you normalize image data? #125

Closed sgmath12 closed 5 years ago

sgmath12 commented 5 years ago

In common setting, we normalize the data something like this (img - img.mean)/(img.std)

But in your code, you didn't divide std.

Is there any reason for that?

wkentaro commented 5 years ago

Because vgg16 doesn't, which is used as a pretrained-model for feature extraction