Authors: Rubèn Tito

Affiliation: Computer Vision Center (CVC)

Email: rperez@cvc.uab.cat

Description: Longformer finetuned on MP-DocVQA using as context the concatenation of the recognized words of all the pages in the document.