Max Pooling Layer Tuning

Hello,

I'm currently working on implementing a CNN for a regression problem, and I've encountered an issue when using the maxPooling2dLayer with a poolSize greater than 1. Here's the relevant part of my code:

Layers = [imageInputLayer([NIV 1 1])  
    convolution2dLayer(5, 15, 'Padding', 'same')
    batchNormalizationLayer
    reluLayer
    maxPooling2dLayer(2)        % (poolSize > 1 leads to an error!!!)
    reluLayer
    dropoutLayer(0.5)
    fullyConnectedLayer(1)
    regressionLayer];

The problem arises when I try to set the poolSize in the maxPooling2dLayer to a value greater than 1. I receive the following errors:

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Error using trainNetwork (line 183)

Invalid network.

Error in CNN (line 280)

net = trainNetwork(XTrain, YTrain, Layers, Options);

Caused by:

Layer 5: Input size mismatch. Size of input to this layer is different from the expected input size.

Inputs to this layer:

from layer 4 (output size 6×1×15)

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

I'm seeking assistance to resolve this issue. Any guidance or suggestions you could provide would be greatly appreciated.

Thank you!

[Matt J]

NIV is missing.

[Jahetbe]

NIV is the 6

Answer 1

[Matt J]

Your input image's spatial dimensions are 6x1. This is not compatible with a 2x2 maxpooling filter. Did you mean to have 2x1 pooling instead? If so,

maxPooling2dLayer([2,1]) 

[Jahetbe]

@Matt J Thanks. It works

[Matt J]

I'm glad, but please Accept-click the answer to indicate so.

[Jahetbe]

Done.