Communication Dans Un Congrès Année : 2026

Zero-Shot Table Extraction in Business Documents: A Unified Benchmark with Error Taxonomy and Ecological Analysis

Résumé

Tables in business documents power analytics and compliance, yet task-specific datasets are costly to build. Practitioners therefore turn to zero-shot vision-language models (VLMs). We study zero-shot realism for table detection (TD) and table structure recognition (TSR) under a unified protocol on DocILE-QUEST and a private STM154 corpus. We report TD with GIoU, Purity, and Completeness, and TSR with TEDS and TEDS-S, evaluating commercial VLMs (GPT-4o, GPT-5-mini), compact detectors, and supervised YOLO/DETR baselines. Zero-shot VLMs are strong for TSR and competitive for TD, while fine-tuned or from-scratch detectors lead when box quality and robustness to clutter matter. We add an automated error taxonomy that isolates actionable failures (missed, merged/split tables, header-body confusions, cell topology). Finally, we quantify emissions, finding a 10 4 gap between the lightest and heaviest systems.

Fichier principal
Vignette du fichier
Thomas_Zero-Shot_Table_Extraction_in_Business_Documents_A_Unified_Benchmark_with_WACV_2026_paper_cvf.pdf (862.62 Ko) Télécharger le fichier
Origine Fichiers produits par l'(les) auteur(s)
Licence

Dates et versions

hal-05616023 , version 1 (07-05-2026)

Licence

Identifiants

Citer

Eliott Thomas, Mickael Coustaty, Aurélie Joseph, Tri-Cong Pham, Gaspar Deloin, et al.. Zero-Shot Table Extraction in Business Documents: A Unified Benchmark with Error Taxonomy and Ecological Analysis. 2026 IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), Mar 2026, Tucson, France. pp.4974-4983, ⟨10.1109/WACV61042.2026.00483⟩. ⟨hal-05616023⟩

Collections

0 Consultations
0 Téléchargements

Altmetric

Partager

  • More