search

sdemyanov / ConvNet

Convolutional Neural Networks for Matlab for classification and segmentation, including Invariang Backpropagation (IBP) and Adversarial Training (AT) algorithms. Trained on GPU, require cuDNN v5.
240 stars 141 forks source link

CNN theory #11

Closed mnaetr closed 9 years ago

mnaetr commented 10 years ago

Hi, i'm using your CNN code. I need some theoretical/matemathical information. I understood that in the convolutional layers each outputmap is obtained by the sum of the convolutions of the different inputmaps with the respective kernel. Is it right? I can't find papers/tutorials that explains this passage in a detailed way. Where can i find it? Can you help me?

rmanor commented 10 years ago

try this: http://cogprints.org/5869/1/cnn_tutorial.pdf or just google for [convolutional neural networks tutorial] , there are many out there.

On Mon, May 26, 2014 at 8:36 PM, mnaetr notifications@github.com wrote:

Hi, i'm using your CNN code. I need some theoretical/matemathical information. I understood that in the convolutional layers each outputmap is obtained by the sum of the convolutions of the different inputmaps with the respective kernel. Is it right? I can't find papers/tutorials that explains this passage in a detailed way. Where can i find it? Can you help me?

— Reply to this email directly or view it on GitHubhttps://github.com/sdemyanov/ConvNet/issues/11 .

mnaetr commented 10 years ago

Ok, thanks. I have another question. If i have a CNN with 2 layers (convolution, subsampling, convolution, subsampling), the first with N1=5 outputmas and the second with N2=3 outputmaps, how many kernel are there in the second convolutional layer? 15 (=N1xN2), is it right? Is this formula valid in general?

sdemyanov commented 10 years ago

Yes, it is right.

On Tue, May 27, 2014 at 6:16 PM, mnaetr notifications@github.com wrote:

Ok, thanks. I have another question. If i have a CNN with 2 layers (convolution, subsampling, convolution, subsampling), the first with N1=5 outputmas and the second with N2=3 outputmaps, how many kernel are there in the second convolutional layer? 15 (=N1xN2), is it right? Is this formula valid in general?

— Reply to this email directly or view it on GitHubhttps://github.com/sdemyanov/ConvNet/issues/11#issuecomment-44246791 .

PhD candidate, Computing and Information Systems, The University of Melbourne.

Sergey Demyanov

mnaetr commented 10 years ago

Thank you very much! Bye