AutoCNN icon indicating copy to clipboard operation
AutoCNN copied to clipboard

Skip Layer - Feature Sizes

Open zeynepalici opened this issue 3 years ago • 2 comments

Hi,

In the related paper, the parameters encoded for a skip layer are the numbers of the feature maps for the two convolutional layers (denoted as F1 and F2, respectively). But you stated them as filter sizes and values of power of 2. I'm trying to figure out the article and implementation. Would you explain to me your point of view about that?

Thank you.

zeynepalici avatar Apr 19 '22 12:04 zeynepalici

Following.

QAQureshi avatar Apr 25 '22 23:04 QAQureshi

Hi,

In the related paper, the parameters encoded for a skip layer are the numbers of the feature maps for the two convolutional layers (denoted as F1 and F2, respectively). But you stated them as filter sizes and values of power of 2. I'm trying to figure out the article and implementation. Would you explain to me your point of view about that?

Thank you.

Here is the code link (implemented by the author of the paper) https://github.com/yn-sun/cnn-ga

QAQureshi avatar Apr 26 '22 00:04 QAQureshi