Why Image classification competition ask 5 candidates answer?
I mean, if Image is about Orange, then (Apple, Car, Airplane, Book, Orange) is right answer.
But if the competition asked 3 candidates, and this machine only answer first 3 -(Apple, Car, Airplane), then it wrong.
So, 5 candidates answer is a lot easier way then 3 candidates. and 3 is ~ 1.
Why 5 candidate answer - of all numbers especially 1- ask?
I spent half hour to research about it.
Firstly, the classifier is not yet very accurate. It can encourage researchers. Secondly, the objects in the picture are not one.