This dataset includes around 600 labeled images from Antiphonale Romanum and Antiphonale Monasticum.
It was created in order to train a model which can recognize groups of square notes for each syllabe and differentiate antiphones from other types of chants.
