I would love to view activation maps from the layers. I implemented the VGG16 model in c++ api. The model is trained in tensorflow framework and the weights have been converted and saved into .wts format as required. Marking the output node to the layer output I want to visualize,
auto out = conv1->getOutput(0);
getting the output of the layer after the inference in the float array[112 * 112 * 64], reshaping it to size, map1. I am not sure if this the correct way. Can anyone please guide.