• @Godnroc
    link
    31 year ago

    Step 1: collect pictures that may contain objects you want to identify Step 2: add some manually identified control pictures Step 3: ask users to identify the object in a sample of the potential pictures using the control pictures for verification