We introduce DocPolarBERT, a layout-aware BERT model for document understanding that eliminates the need for absolute 2D positional embeddings. We extend self-attention to take into account text block positions in relative polar coordinate system rather than the Cartesian one. Despite being pre-trained on a dataset more than six times smaller than the widely used IIT-CDIP corpus, DocPolarBERT achieves state-of-the-art results. These results demonstrate that a carefully designed attention mechanism can compensate for reduced pre-training data, offering an efficient and effective alternative for document understanding.
@article{arxiv.2507.08606,
title = {DocPolarBERT: A Pre-trained Model for Document Understanding with Relative Polar Coordinate Encoding of Layout Structures},
author = {Benno Uthayasooriyar and Antoine Ly and Franck Vermet and Caio Corro},
journal= {arXiv preprint arXiv:2507.08606},
year = {2026}
}