Semi-translated dataset of CUB-200-2011 into Indonesian. This dataset contains thousands
of image-text annotation pairs of 200 subcategories belonging to birds. The natural
language descriptions are collected through the Amazon Mechanical Turk (AMT) platform and
are required at least 10 words, without any information on subcategories and actions.
