An error when run Unet Model Deep learning

Dear all,

I was develop my network thru apps design. then generate code. Like below.

But I got error.

clc
clear all
close all
%testDataimages
DATASetDir = fullfile('C:\Users\Akmal\Desktop\NEW 3D U NET 128X128');
IMAGEDir = fullfile(DATASetDir,'ImagesTr');
volReader = @(x) matRead(x);
volds = imageDatastore(IMAGEDir, ...
    'FileExtensions','.mat','ReadFcn',volReader);
% labelReader = @(x) matread(x);
matFileDir = fullfile('C:\Users\Akmal\Desktop\NEW 3D U NET 128X128\LabelsTr');
classNames = ["background", "tumor"];
pixelLabelID = [0 1];
% pxds = (LabelDirr,classNames,pixelLabelID, ...
%     'FileExtensions','.mat','ReadFcn',labelReader);
pxds = pixelLabelDatastore(matFileDir,classNames,pixelLabelID, ...
    'FileExtensions','.mat','ReadFcn',@matRead);
ds = pixelLabelImageDatastore(volds,pxds);
volume = preview(volds);
label = preview(pxds);
patchSize = [128 128 64];
patchPerImage = 16;
miniBatchSize = 8;
patchds = randomPatchExtractionDatastore(volds,pxds,patchSize, ...
    'PatchesPerImage',patchPerImage);
patchds.MiniBatchSize = miniBatchSize;
dsTrain = transform(patchds,@augment3dPatch);
volLocVal = fullfile('C:\Users\Akmal\Desktop\NEW 3D U NET 128X128\imagesVal');
voldsVal = imageDatastore(volLocVal, ...
    'FileExtensions','.mat','ReadFcn',volReader);
lblLocVal = fullfile('C:\Users\Akmal\Desktop\NEW 3D U NET 128X128\labelsVal');
pxdsVal = pixelLabelDatastore(lblLocVal,classNames,pixelLabelID, ...
    'FileExtensions','.mat','ReadFcn',volReader);
dsVal = randomPatchExtractionDatastore(voldsVal,pxdsVal,patchSize, ...
    'PatchesPerImage',patchPerImage);
dsVal.MiniBatchSize = miniBatchSize;
lgraph = layerGraph();
tempLayers = [
    image3dInputLayer([128 128 64 1],"Name","image3dinput")
    convolution3dLayer([3 3 3],64,"Name","Encoder-Stage-1-Conv-1","Padding","same")
    reluLayer("Name","Encoder-Stage-1-ReLU-1_1")
    %126x126
    convolution3dLayer([3 3 3],64,"Name","Encoder-Stage-1-Conv-2","Padding","same")
    reluLayer("Name","Encoder-Stage-1-ReLU-1_2")];
%124x124
lgraph = addLayers(lgraph,tempLayers);
tempLayers = [
    maxPooling3dLayer([2 2 2],"Name","Encoder-Stage-1-MaxPool_1","Padding","same")
    %62x62
    convolution3dLayer([3 3 3],128,"Name","Encoder-Stage-1-MaxPool_2","Padding","same")
    reluLayer("Name","Encoder-Stage-2-ReLU-1")
    %60x60
    convolution3dLayer([3 3 3],128,"Name","Encoder-Stage-2-Conv-2","Padding","same")
    reluLayer("Name","Encoder-Stage-2-ReLU-2")];
%58x58
lgraph = addLayers(lgraph,tempLayers);
tempLayers = [
    maxPooling3dLayer([2 2 2],"Name","Encoder-Stage-2-MaxPool","Padding","same")
    %29x29
    convolution3dLayer([3 3 3],256,"Name","Encoder-Stage-3-Conv-1","Padding","same")
    reluLayer("Name","Encoder-Stage-3-ReLU-1")
    %27x27
    convolution3dLayer([3 3 3],256,"Name","conv3d","Padding","same")
    reluLayer("Name","Encoder-Stage-3-ReLU-2")];
%25x25
lgraph = addLayers(lgraph,tempLayers);
tempLayers = [
    dropoutLayer(0.5,"Name","Encoder-Stage-3-DropOut")
    maxPooling3dLayer([2 2 2],"Name","Encoder-Stage-3-MaxPool_1","Padding","same")
    %12x12
    convolution3dLayer([3 3 3],512,"Name","Bridge-Conv-1","Padding","same")
    reluLayer("Name","Bridge-ReLU-1")
    %10x10
    convolution3dLayer([3 3 3],512,"Name","Bridge-Conv-2","Padding","same")
    reluLayer("Name","Bridge-ReLU-2")
    %8x8
    dropoutLayer(0.5,"Name","Bridge-DropOut")
    transposedConv3dLayer([2 2 2],256,"Name","Decoder-Stage-1-UpConv","Cropping","same")
    reluLayer("Name","Decoder-Stage-1-UpReLU")];
lgraph = addLayers(lgraph,tempLayers);
tempLayers = [
    depthConcatenationLayer(2,"Name","Decoder-Stage-1-DepthConcatenation")
    convolution3dLayer([3 3 3],256,"Name","Decoder-Stage-1-Conv-1","Padding","same")
    reluLayer("Name","Decoder-Stage-1-ReLU-1")
    convolution3dLayer([3 3 3],256,"Name","Decoder-Stage-1-Conv-2","Padding","same")
    reluLayer("Name","Decoder-Stage-1-ReLU-2")
    transposedConv3dLayer([2 2 2],128,"Name","Decoder-Stage-2-UpConv","Cropping","same")
    reluLayer("Name","Decoder-Stage-2-UpReLU")];
lgraph = addLayers(lgraph,tempLayers);
tempLayers = [
    depthConcatenationLayer(2,"Name","Decoder-Stage-2-DepthConcatenation")
    convolution3dLayer([3 3 3],128,"Name","Decoder-Stage-2-Conv-1","Padding","same")
    reluLayer("Name","Decoder-Stage-2-ReLU-1")
    convolution3dLayer([3 3 3],128,"Name","Decoder-Stage-2-Conv-2","Padding","same")
    reluLayer("Name","Decoder-Stage-2-ReLU-2")
    transposedConv3dLayer([2 2 2],64,"Name","Decoder-Stage-3-UpConv","Cropping","same")
    reluLayer("Name","Decoder-Stage-3-UpReLU")];
lgraph = addLayers(lgraph,tempLayers);
tempLayers = [
    depthConcatenationLayer(2,"Name","Decoder-Stage-3-DepthConcatenation")
    convolution3dLayer([3 3 3],64,"Name","Decoder-Stage-3-Conv-1","Padding","same")
    reluLayer("Name","Decoder-Stage-3-ReLU-1")
    convolution3dLayer([3 3 3],64,"Name","Decoder-Stage-3-Conv-2","Padding","same")
    reluLayer("Name","Decoder-Stage-3-ReLU-2")
    convolution3dLayer([2 2 2],32,"Name","Final-ConvolutionLayer","Padding","same")
    softmaxLayer("Name","softmax")
    pixelClassificationLayer("Name","pixel-class")];
lgraph = addLayers(lgraph,tempLayers);
% clean up helper variable
clear tempLayers;
lgraph = connectLayers(lgraph,"Encoder-Stage-1-ReLU-1_2","Encoder-Stage-1-MaxPool_1");
lgraph = connectLayers(lgraph,"Encoder-Stage-1-ReLU-1_2","Decoder-Stage-3-DepthConcatenation/in2");
lgraph = connectLayers(lgraph,"Encoder-Stage-2-ReLU-2","Encoder-Stage-2-MaxPool");
lgraph = connectLayers(lgraph,"Encoder-Stage-2-ReLU-2","Decoder-Stage-2-DepthConcatenation/in2");
lgraph = connectLayers(lgraph,"Encoder-Stage-3-ReLU-2","Encoder-Stage-3-DropOut");
lgraph = connectLayers(lgraph,"Encoder-Stage-3-ReLU-2","Decoder-Stage-1-DepthConcatenation/in1");
lgraph = connectLayers(lgraph,"Decoder-Stage-1-UpReLU","Decoder-Stage-1-DepthConcatenation/in2");
lgraph = connectLayers(lgraph,"Decoder-Stage-2-UpReLU","Decoder-Stage-2-DepthConcatenation/in1");
lgraph = connectLayers(lgraph,"Decoder-Stage-3-UpReLU","Decoder-Stage-3-DepthConcatenation/in1");
plot(lgraph);
maxEpochs = 10;
options = trainingOptions('adam', ...
    'MaxEpochs',maxEpochs, ...
    'InitialLearnRate',1e-3, ...
    'LearnRateSchedule','piecewise', ...
    'LearnRateDropPeriod',5, ...
    'LearnRateDropFactor',0.97, ...
    'ValidationData',dsVal, ...
    'ValidationFrequency',50, ...
    'Plots','training-progress', ...
    'Verbose',false, ...
    'MiniBatchSize',miniBatchSize);
doTraining = true;
if doTraining
    modelDateTime = datestr(now,'dd-mmm-yyyy-HH-MM-SS');
    [net,info] = trainNetwork(dsTrain,lgraph,options);
    save(['trained3DUNet-' modelDateTime '-Epoch-' num2str(maxEpochs) '.mat'],'net');
else
    load('trained3DVNet-07-Jun-2022-13-45-30-Epoch-250.mat');
end

ERROR

Error using trainNetwork

Invalid training data. The output size (32) of the last layer does not match the number of

classes of the responses (2).

Error in basicunetmodel3D (line 170)

[net,info] = trainNetwork(dsTrain,lgraph,options);

Anyone can help me?

[KSSV]

It is a problem with input and target. What are the dimensions of input and target?How they are?

[mohd akmal masud]

My input size is 130x130x64

Answer 1

[Ben]

You should set the number of filters on the final convolution layer to match the number of pixel classes for your problem. Tt looks like you have 2 classes, so I would swap convolution3dLayer([2 2 2],32,"Name","Final-ConvolutionLayer","Padding","same") for convolution3dLayer([2 2 2],2,"Name","Final-ConvolutionLayer","Padding","same").

You might also be able to use unetLayers to construct your network, which allows you to specify the numClasses as an input.