ASFF icon indicating copy to clipboard operation
ASFF copied to clipboard

Published 20 hours ago • verified publisher icon GOATmessi7

Why do you use Convolutional layers without Batch-normalization in RFBblock?

Open AlexeyAB opened this issue 4 years ago • 1 comments

@ruinmessi Hi,

Why do you use Convolutional layers without Batch-normalization in RFBblock? https://github.com/ruinmessi/ASFF/blob/55b6637b21d69509ae73100f9db19dc98acb419d/models/network_blocks.py#L138-L159

AlexeyAB avatar Dec 16 '19 17:12 AlexeyAB

@AlexeyAB Hi, I just do this for simplicity. I think the essential part of rfb is the distribution of RFs, not BN. But I haven't done any experiments to verify that.

GOATmessi7 avatar Dec 17 '19 01:12 GOATmessi7