# Test data from ImageNet

Here's the description about the data usage for ILSVRC 2016 of ImageNet. I've interpreted it with the table as follows,

Data          Image        Label    Changed    Release_Time
-----------------------------------------------------------------
training      1,200,000    1,000    no         always public
validation       50,000    1,000    some       upon registration
test            100,000        0    some       final evaluation


My question is specific about the test data. Is it chosen from the same set of 150,000 images since ILSVRC 2012, although randomly ? If so, most of the test data has been made public before the final evaluation phase of the challenge, how the result could remain fair enough ?