Abstract
The ARTigo dataset, generated from over 10 million annotations, is a product of a citizen science project developed by the Institute of Art History and the Institute of Computer Science at LMU Munich. The project leverages Games with a Purpose (GWAPs) to foster a playful environment for tagging artworks. In these GWAPs, two anonymous players are given an image to annotate with textual or visual descriptors within a limited time frame. The annotations serve to improve the accessibility of art-historical images and offer vast research potential well beyond their utility as training datasets for Computer Vision (CV) algorithms.