@InProceedings{PauRiba2019,
author="Pau Riba
and Anjan Dutta
and Lutz Goldmann
and Alicia Fornes
and Oriol Ramos Terrades
and Josep Llados",
title="Table Detection in Invoice Documents by Graph Neural Networks",
booktitle="15th International Conference on Document Analysis and Recognition",
year="2019",
pages="122--127",
abstract="Tabular structures in documents offer a complementary dimension to the raw textual data, representing logical or quantitative relationships among pieces of information. In digital mail room applications, where a large amount ofadministrative documents must be processed with reasonable accuracy, the detection and interpretation of tables is crucial. Table recognition has gained interest in document image analysis, in particular in unconstrained formats (absence of rule lines, unknown information of rows and columns). In this work, we propose a graph-based approach for detecting tables in document images. Instead of using the raw content (recognized text), we make use of the location, context and content type, thus it is purely a structure perception approach, not dependent on the language and the quality of the textreading. Our framework makes use of Graph Neural Networks (GNNs) in order to describe the local repetitive structural information of tables in invoice documents. Our proposed model has been experimentally validated in two invoice datasets and achieved encouraging results. Additionally, due to the scarcityof benchmark datasets for this task, we have contributed to the community a novel dataset derived from the RVL-CDIP invoice data. It will be publicly released to facilitate future research.",
optnote="DAG; 600.140; 601.302; 602.167; 600.121; 600.141",
optnote="exported from refbase (http://refbase.cvc.uab.es/show.php?record=3355), last updated on Mon, 30 Nov 2020 10:46:35 +0100",
doi="10.1109/ICDAR.2019.00028",
opturl="https://ieeexplore.ieee.org/document/8978079",
file=":http://refbase.cvc.uab.es/files/RDG2019.pdf:PDF"
}