I have a problem with understanding the actual architecture of this network. Usually, LeCun et al have used different weights for the connections from different feature maps of a previous layer (something that looks like 3D kernel). Therefore, the number of weights of a convolution layer (assuming full map of connections) is kernelHeight*kernelWidth*numFeatMapsLayer(k)*numFeatMapsLayer(k-1). I am not sure, but it seems that there are kernelHeight*kernelWidth*numFeatMapsLayer(k) different weights used in this program. Does it means that the connections from different feature maps of a previous layer to a particular feature map of the next layer have the same weights? Or maybe, I misunderstand something?
Very useful tool!
However, in the latest version 20 Jun 2010, the maximum allowed m-file-name for class files (scripts and functions are ok) is 40 characters! For longer file names, fdep will produce an obscure error message:
In an assignment A(:) = B, the number of elements in A and B must be the same.
It took me some time to figure this out... Unfortunately I have no solution other than to use shorter class names at the moment. This does not seem to be related to depfun.
Thanks for providing these files to the comunity !
However, did you check the othogonality of the functions created?
It seems to me that something's wrong there.
I generated 2 polynoms like it is shown in the doc :
z(idx) = zernfun(3,1,r(idx),theta(idx));
z2(idx) = zernfun(4,0,r(idx),theta(idx));
When I check the orthogonality by simply doing :
the result is not 0.
Am I doing someting wrong or is there a pb here?
Thanks in adavnce