I face this error "Caused by: Layer 17: Input size mismatch. Size of input to this layer is different from the expected input size. Inputs to this layer: from layer 16 (1×1×2 output) "

layers = [
    imageInputLayer([8 1 1])
    
    convolution2dLayer([4 1],8,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer([2 1],'Stride',2)
    
     convolution2dLayer([4 1],16,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer([2 1],'Stride',2)
    
    convolution2dLayer([4 1],4,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer([2 1],'Stride',2)
   
    convolution2dLayer([4 1],2,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    maxPooling2dLayer([2 1],'Stride',2)
    
    
    convolution2dLayer([4 1],1,'Padding','same')
    batchNormalizationLayer
    reluLayer
    
    fullyConnectedLayer(2)
    softmaxLayer
    classificationLayer];    

      
options = trainingOptions('sgdm',...
    'ExecutionEnvironment', 'cpu',...
    'InitialLearnRate',0.005,...
    'MaxEpochs',30, ...
    'Verbose',false,...
    'ValidationData',{XValidation,YValidation}, ...
    'ValidationFrequency',10, ...
    'MiniBatchSize',30, ...
    'Plots','training-progress');

net = trainNetwork(Xtrain,Ytrain,layers,options);   %

Answer 1

[Garmit Pant]

Hello Saddam

The error you have encountered is due to a size mismatch between the expected input size of the fourth max pooling layer and the actual size of the input of the layer.

On analysing the network using “analyzeNetwork” function, we can see that the output size after layer 14, ‘conv_4’, is 1x1x2x1. The size of the pool operation on layer 17, ‘maxpool_4’, is 2x1. Since the size of the pooling dimensions should be larger or equal to the pool size, the error you are encountering is occurring. You can use the following code snippet to check your network:

analyzeNetwork(layers) 

Output:

This can be rectified by removing the fourth max pooling layer.

For further understanding, kindly refer to the following MathWorks Documentation:

• Refer to the ‘Properties’ section to understand how to correctly set up the max pooling layer: https://www.mathworks.com/help/deeplearning/ref/nnet.cnn.layer.maxpooling2dlayer.html
• Refer to the ‘Algorithms’ section to understand the input and corresponding output sizes of the 2D convolutional layer: https://www.mathworks.com/help/deeplearning/ref/nnet.cnn.layer.convolution2dlayer.html

I hope you find the above explanation and suggestions useful!