Giters

ikhlestov / vision_networks

Repo about neural networks for images handling

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Problem regarding the number of features generated at the initial convolution layer

curiositywan opened this issue · comments

commented

In Densenet paper, it mentions that for non-BC architectures, the number of the output features for the initial conv layer should be 16, and for BC architectures, it's twice as the growth rate. However, in your implementation, the number is always twice as the growth rate.

commented

Really I adopted code from initial repo(link to code line), where it seems to me features from the initial convolution always generated as grows_rate*2. If it's incorrect and may cause problem - please, fix me.