English

Chargrid: Towards Understanding 2D Documents

Computation and Language 2018-09-25 v1 Computer Vision and Pattern Recognition Machine Learning Neural and Evolutionary Computing

Abstract

We introduce a novel type of text representation that preserves the 2D layout of a document. This is achieved by encoding each document page as a two-dimensional grid of characters. Based on this representation, we present a generic document understanding pipeline for structured documents. This pipeline makes use of a fully convolutional encoder-decoder network that predicts a segmentation mask and bounding boxes. We demonstrate its capabilities on an information extraction task from invoices and show that it significantly outperforms approaches based on sequential text or document images.

Keywords

Cite

@article{arxiv.1809.08799,
  title  = {Chargrid: Towards Understanding 2D Documents},
  author = {Anoop Raveendra Katti and Christian Reisswig and Cordula Guder and Sebastian Brarda and Steffen Bickel and Johannes Höhne and Jean Baptiste Faddoul},
  journal= {arXiv preprint arXiv:1809.08799},
  year   = {2018}
}

Comments

To be published at EMNLP 2018

R2 v1 2026-06-23T04:15:59.584Z