Revising FUNSD dataset for key-value detection in document images
This work addresses data quality issues for researchers in document information extraction, but it is incremental as it focuses on dataset revision and minor model improvements.
The authors identified labeling inconsistencies in the FUNSD dataset for key-value detection in document images and revised it to improve applicability, reporting baseline results with a UNet model and an improved version using Channel-Invariant Deformable Convolution.
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.