Document Recognition and Document Intelligence

Overview

We study document understanding across scanned and born-digital content, covering layout parsing, entity linking, table/chart understanding, and robust reasoning with multimodal cues.

Outcomes

• 

  GraphLayoutLM

  Enhancing visually-rich document understanding via layout structure modeling. MM 2023.

  2023 Document AI Graph
• 

  Hypergraph for Document Entity Recognition

  Hypergraph-based understanding for document semantic entity recognition. ACL 2024.

  2024 ACL Hypergraph