EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

A document is worth a structured record: Principled inductive bias design for document recognition

Benjamin Meyer, Lukas Tuggener, Sascha Hänzi, Daniel Schmid, Erdal Ayfer, Benjamin F. Grewe, Ahmed Abdulkadir, Thilo Stadelmann

International Journal on Document Analysis and Recognition (IJDAR) · 23.07.2026

DOI: 10.1007/s10032-026-00602-9 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal on Document Analysis and Recognition (IJDAR), online erschienen am 23.07.2026. Verfasst von Benjamin Meyer und 7 weiteren Autor:innen. Die Arbeit nennt 94 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Innosuisse - Schweizerische Agentur für Innovationsförderung, ZHAW School of Engineering.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Many document types use intrinsic, convention-driven structures that serve to encode precise and structured information, such as the conventions governing engineering drawings. However, many state-of-the-art approaches treat document recognition as a mere computer vision problem, neglecting these underlying document-type-specific structural properties, making them dependent on sub-optimal heuristic post-processing and rendering many less frequent or more complicated document types inaccessible to modern document recognition. We suggest a novel perspective that frames document recognition as a transcription task from a document to a record. This implies a natural grouping of documents based on the intrinsic structure inherent in their transcription, where related document types can be treated (and learned) similarly. We propose a method to design structure-specific relational inductive biases for the underlying machine-learned end-to-end document recognition systems, and a respective base transformer architecture that we successfully adapt to different structures. We demonstrate the effectiveness of the so-found inductive biases in extensive experiments with progressively complex record structures from monophonic sheet music, shape drawings, and simplified engineering drawings. By integrating an inductive bias for unrestricted graph structures, we train the first-ever successful end-to-end model to transcribe mechanical engineering drawings to their inherently interlinked information. Our approach is relevant to inform the design of document recognition systems for document types that are less well understood than standard OCR, OMR, etc., and serves as a guide to unify the design of future document foundation models.

Förderung

  • Innosuisse - Schweizerische Agentur für Innovationsförderung
  • ZHAW School of Engineering

Zitieren

APA

Meyer, B., Tuggener, L., Hänzi, S., Schmid, D., Ayfer, E., Grewe, B. F., Abdulkadir, A., & Stadelmann, T. (2026). A document is worth a structured record: Principled inductive bias design for document recognition. International Journal on Document Analysis and Recognition (IJDAR). https://doi.org/10.1007/s10032-026-00602-9

Vancouver

Meyer B, Tuggener L, Hänzi S, Schmid D, Ayfer E, Grewe BF, et al. A document is worth a structured record: Principled inductive bias design for document recognition. International Journal on Document Analysis and Recognition (IJDAR). 2026. doi:10.1007/s10032-026-00602-9

BibTeX
@article{Meyer2026,
  author = {Meyer, Benjamin and Tuggener, Lukas and Hänzi, Sascha and Schmid, Daniel and Ayfer, Erdal and Grewe, Benjamin F. and Abdulkadir, Ahmed and Stadelmann, Thilo},
  title = {A document is worth a structured record: Principled inductive bias design for document recognition},
  journal = {International Journal on Document Analysis and Recognition (IJDAR)},
  year = {2026},
  doi = {10.1007/s10032-026-00602-9}
}
Kostenlos starten →