Revising FUNSD dataset for key-value detection in document images

TL;DR

This paper revises the FUNSD dataset for key-value detection in document images by addressing labeling inconsistencies and demonstrates baseline and improved models for key-value extraction.

Contribution

It identifies labeling issues in FUNSD and provides a revised version along with baseline and enhanced models for key-value detection.

Findings

Revised FUNSD dataset with corrected labels.

Baseline UNet model for key-value detection.

Improved UNet with Channel-Invariant Deformable Convolution.

Abstract

FUNSD is one of the limited publicly available datasets for information extraction from document im-ages. The information in the FUNSD dataset is defined by text areas of four categories ("key", "value", "header", "other", and "background") and connectivity between areas as key-value relations. In-specting FUNSD, we found several inconsistency in labeling, which impeded its applicability to thekey-value extraction problem. In this report, we described some labeling issues in FUNSD and therevision we made to the dataset. We also reported our implementation of for key-value detection onFUNSD using a UNet model as baseline results and an improved UNet model with Channel-InvariantDeformable Convolution.