yuhanlolo/deep-learning-faces

Automatically exported from code.google.com/p/deep-learning-faces

deep-learning-faces's Issues

compiling error on make mexf=para

I encountered an error on compiling step.
I'm using Cuda 5.5, on ubuntu 12.04 64 bit

I'd followed the steps written in README,

Once I type "make mexf="./deep_nn/mexcuConvNNoo.mex 
./deep_nn/mexcuConvNNooFF.mex", 

it returns the error:
make: /pkgs/matlab-80/bin/mexext: Command not found
make: *** No rule to make target 'deep_nn/mexcuConvNNoo.', needed by 'all'. 
Stop.

I've tried to figure it myself, but it seems to be difficult to me.
Sorry for that I have lack of knowledge on this, I'll doing on my best.

Sincerely
Tae-Ho Kim

Original issue reported on code.google.com by [email protected] on 23 Nov 2013 at 11:59

No data set available.

dataset has been removed from Kaggle website

Original issue reported on code.google.com by [email protected] on 7 Oct 2014 at 5:46

Invalid MEX-file error parameters.

Invalid MEX-file 
'D:\DEEPLearning\2013\SVM_SOFTMAX\matlab\mexcuConvNNoo.mexw64': error 
parameters.
Error in myclassify_conv_nn_softmax (line 174)
        [~, model.theta] = mexcuConvNNoo( single(ww), params, model.callback_name);
        in code write like this:
#define NARGIN 3
#define IN_W pRhs[0]
#define IN_params pRhs[1]
#define IN_callback  pRhs[2]

#define NARGOUT 1
#define OUT_f pLhs[0]
#define OUT_W pLhs[1]
#define OUT_dW pLhs[1]
and which is different from the beginning explanation:function [f (W or dW)] = 
mexcuConvNNoo( W, params, Dim, Callback).

Original issue reported on code.google.com by [email protected] on 12 Aug 2013 at 8:25