Error using nnet.cnn.L​ayerGraph>​iThrowErro​rIfLayerHa​sMultipleI​nputs Layer 'inception_3a-output' has multiple inputs. Specify which input of layer 'inception_3a-output' to use.

clc

clear all;

close all;

%Open file in MATLAB

outputFolder=fullfile('recycle101');

rootFolder=fullfile(outputFolder,'project');

categories={'aluminiumcan','petbottle','drinkcartonbox'};

%Separate dataset into training and validation in the ratio of 70:30

imds=imageDatastore(fullfile(rootFolder,categories),'LabelSource','foldernames');

tbl=countEachLabel(imds)

minSetCount=min(tbl{:,2});

[imdsTrain,imdsValidation] = splitEachLabel(imds,0.7,'randomized');

%Count the total datasets in each file

countEachLabel(imds);

%Randomly choose file for aluminium can, PET bottles, and drink carton box

AluminiumCan=find(imds.Labels=='aluminiumcan',1);

PETBottles=find(imds.Labels=='petbottle',1);

DrinkCartonBox=find(imds.Labels=='drinkcartonbox',1);

%Plot image that was pick randomly

figure

subplot(2,2,1);

imshow(readimage(imds,AluminiumCan));

subplot(2,2,2);

imshow(readimage(imds,PETBottles));

subplot(2,2,3);

imshow(readimage(imds,DrinkCartonBox));

%Load pre-trained network

net=googlenet;

analyzeNetwork(net)

lys = net.Layers;

lys(end-3:end)

numClasses = numel(categories(imdsTrain.Labels));

lgraph = layerGraph(lys);

%Replace the classification layers for new task

newFCLayer = fullyConnectedLayer(3,'Name','new_fc','WeightLearnRateFactor',10,'BiasLearnRateFactor',10);

lgraph = replaceLayer(lgraph,'loss3-classifier',newFCLayer);

newClassLayer = classificationLayer('Name','new_classoutput');

lgraph = replaceLayer(lgraph,'output',newClassLayer);

analyzeNetwork(lgraph)

lgraph = connectLayers(lgraph,'inception_3a-relu_5x5','inception_3a-output');

lgraph = connectLayers(lgraph,'inception_3b-output','pool3-3x3_s2');

%Resize the image and train the network

imageSize=net.Layers(1).InputSize;

augmentedTrainingSet=augmentedImageDatastore(imageSize,...

imdsTrain,'ColorPreprocessing','gray2rgb');

augmentedValidateSet=augmentedImageDatastore(imageSize,...

imdsValidation,'ColorPreprocessing','gray2rgb');

options = trainingOptions('sgdm', ...

'MiniBatchSize',4, ...

'MaxEpochs',8, ...

'InitialLearnRate',1e-4, ...

'Shuffle','every-epoch', ...

'ValidationData',augmentedValidateSet, ...

'ValidationFrequency',3, ...

'Verbose',false, ...

'ExecutionEnvironment','cpu', ...

'Plots','training-progress');

trainedNet = trainNetwork(augmentedTrainingSet,lgraph,options);

Answer 1

[Ranjeet]

Hi Tan,

You are extracting layers from ‘GoogleNet’ in the following code and then using the extracted layer to form a ‘layerGraph’ and then further add/replace other layers like fully connection layer and classification layer.

%Load pre-trained network
net=googlenet;
analyzeNetwork(net)
lys = net.Layers;
lys(end-3:end)
numClasses = numel(categories(imdsTrain.Labels));
lgraph = layerGraph(lys);

However, ‘Googlenet’ has multiple parallel connections from a single layer throughout the network. When you are create layerGraph from net.Layers, connection information in ‘net.Connections’ is not being used. ‘layerGraph’ can directly be created from network by ‘lgraph = layerGraph(net)’ so that connection information is also there. You may use the following code:

net=googlenet;
analyzeNetwork(net)

numClasses = 5;
lgraph = layerGraph(net);

% check the layer connected to fully connected layer]
inputSizeFc = lgraph.Layers(142).InputSize;

%Replace the classification layers for new task
newFCLayer = fullyConnectedLayer(numClasses,'Name','new_fc','WeightLearnRateFactor',10,'BiasLearnRateFactor',10, ...
    'Bias',ones(numClasses, 1), 'Weights',ones(numClasses, inputSizeFc));
lgraph = replaceLayer(lgraph,'loss3-classifier',newFCLayer);
newClassLayer = classificationLayer('Name','new_classoutput');
lgraph = replaceLayer(lgraph,'output',newClassLayer);

analyzeNetwork(lgraph);