Hi, great paper. I wanted to ask if there was any reason why performance on the unseen set was not reported or what is the performance from the current method on the unseen test set provided by the challenge?
This is to allow for comparison with other published work. Thank you.