English

DocPolarBERT: A Pre-trained Model for Document Understanding with Relative Polar Coordinate Encoding of Layout Structures

Computation and Language 2026-01-23 v4

Abstract

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.

Keywords

Cite

@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}
}

Comments

EACL 2026 (main)

R2 v1 2026-07-01T03:56:38.577Z