Search code examples
cntk

Do we have to transform input before evaluation using CNTK C++ Library?

I've trained cntk model and now I want to evaluate it in my c++ code using CNTKLibrary. Configuration for training had this line:

featNorm = features - 128
p1 = MyLayer (featNorm, 32, 0.1557/256)

Does this mean that I have to deduct 128 from pixel values when I create NDArrayView to pass as input data for evaluation, or it is written somewhere in model and CNTKLibrary will normalize input data itself?

Solution

  • Because the normalization was done as part of the graph, you don't need to subtract 128 in your C++ code.