EUVIMEDEUVIMED entdecken
Crossref

P‐110: Chain‐of‐Thought Enhanced Vision‐Language Model for Key Information Extraction in Logistics Documents of the Display Industry

Pan Zhong, Guangwei Huang, Bin Lv, Xingqun Jiang

SID Symposium Digest of Technical Papers · 2026-06 · 57 (1) · 2115-2118

DOI: 10.1002/sdtp.70001_575 ↗

Kurz gefasst

Fachartikel in der Zeitschrift SID Symposium Digest of Technical Papers, online erschienen am 11.09.2026. Verfasst von Pan Zhong und 3 weiteren Autor:innen. Die Arbeit nennt 13 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Tackling the challenges of heterogeneous document layouts and implicit semantics in cross‐border logistics of the display industry, this paper proposes a Chain‐of‐Thought (CoT) enhanced Vision‐Language Model framework for key information extraction. The proposed approach leverages a self‐constructed KIE‐CoT dataset and reinforcement learning to build explicit reasoning paths for field‐level understanding. Experimental results show that the model achieves 98.8% field‐level accuracy and 87.9% document‐level accuracy, with the latter representing a 3.4% improvement over strong baselines. These results demonstrate the framework's effectiveness in document understanding and its potential to provide essential technological support for the digital development of the logistics system in the display industry.

Zitieren

APA

Zhong, P., Huang, G., Lv, B., & Jiang, X. (2026). P‐110: Chain‐of‐Thought Enhanced Vision‐Language Model for Key Information Extraction in Logistics Documents of the Display Industry. SID Symposium Digest of Technical Papers, 57(1), 2115-2118. https://doi.org/10.1002/sdtp.70001_575

Vancouver

Zhong P, Huang G, Lv B, Jiang X. P‐110: Chain‐of‐Thought Enhanced Vision‐Language Model for Key Information Extraction in Logistics Documents of the Display Industry. SID Symposium Digest of Technical Papers. 2026;57(1):2115-2118. doi:10.1002/sdtp.70001_575

BibTeX
@article{Zhong2026,
  author = {Zhong, Pan and Huang, Guangwei and Lv, Bin and Jiang, Xingqun},
  title = {P‐110: Chain‐of‐Thought Enhanced Vision‐Language Model for Key Information Extraction in Logistics Documents of the Display Industry},
  journal = {SID Symposium Digest of Technical Papers},
  year = {2026},
  volume = {57},
  number = {1},
  pages = {2115--2118},
  doi = {10.1002/sdtp.70001_575}
}
Kostenlos starten →