I am using hdf5 layer that has data and label.
layer {
name: "data"
type: "HDF5Data"
top: "data"
top: "label"
include {
phase: TEST
}
hdf5_data_param {
source: "./list.txt"
batch_size: 8
shuffle: true
}
}
During the testing phase, it will load 8 images from the testing set and feed the network. I would like to print the mean value of each image during the testing phase. Is it possible to use it in CAFFE? Which layer should I used?