CNNs: Does classification use the same normalization as input layer?

Dear colleagues,

Does anyone knows if the "classify" function to classify data using a trained deep learning network uses the same normalization as the imageinputlayer?

For example,

if I use: "imageInputLayer(inputSize, "Normalization", "zscore", "NormalizationDimension", "element")

does the

[pred,scores] = classify(trainedNet, Test_imds)

applies the same normalization to the "Test_imds" during the classification?

If not, how can I apply the exact same normalization to "Test_imds"?

THANK YOU IN ADVANCE!!