Why CNN Regression trainnetwork input output be the same 2-D metrics

I want to use same size 2D Input Output data to build a denoising CNN model just like Resnet

But net = trainNetwork(X,X,layers,options) always sending error:

Invalid training data. X and Y must have the same number of observations.

And I do not understand how to fix it . Here is a sample code below, Thanks for your help

layers = [...
  imageInputLayer([25 25 1])
  convolution2dLayer([4 3],12)
  reluLayer
  crossChannelNormalizationLayer(4)
  maxPooling2dLayer(2,'Stride',2)
  convolution2dLayer(5,16)
  reluLayer
  crossChannelNormalizationLayer(4)
  maxPooling2dLayer(2,'Stride',2)
  fullyConnectedLayer(256)
  reluLayer
  fullyConnectedLayer(25*25)%Change the dimensions to match the outputs.
  regressionLayer;
  ];
%create training option
options = trainingOptions('sgdm','InitialLearnRate',0.001, ...
  'MaxEpochs',15);
%create network
      X(:,:,:,1) = rand(25);
      Y=randn(1,25,25,1);
net = trainNetwork(X,Y,layers,options)

Answer 1

[Siraj]

Hi,

It is my understanding that you want to train a model that gives the output of same dimension that of the input. For this you will have to tune your architecture, Currently the output dimensions of your network are 1X1X625 because the fully connected layer always gives a 1D output, also the error that you are getting is because the Y has incorrect dimensions. The responses, Y, can be a NxR matrix, where N = observations, R = number of responses (or output of the network).

Refer to Input Arguments of trainNetwork.

Hope it helps