FUNSD Dataset for Document Understanding

The FUNSD dataset is a resource for form understanding in noisy scanned documents, created by JetsonEarth. It was last updated on July 30, 2022.

Use Cases

• Train a model for semantic entity labeling on form documents.
• Develop a layout analysis system to understand relationships between form fields.
• Fine-tune a model for optical character recognition and text extraction from forms.
• Build a system for information extraction from structured and semi-structured documents.

Strengths

• Dataset is associated with the FUNSD benchmark for form understanding.
• Last update was recorded on July 30, 2022, providing a specific timestamp.

Limitations

• The dataset's size, row count, column structure, and specific file formats are unknown.
• The license terms are unspecified, which may restrict usage.
• The geographic scope and temporal coverage of the data are not defined.

Provenance

Source

JetsonEarth

Freshness

Last updated on 2022-07-30.