Hello,
I was wondering how I can apply the convolution 2D operation, as performed in learning networks, without the training phase.
I just would like to replicate the convolution operation.
For example, if I want to map an m x n x u tensor into a m x n x v tensor, how can I perform such operation?
I guess I could apply the convn operator, but I do not know how to properly use it.
Thanks
Simone
Best Answer