LaTeX2Layout: High-Fidelity, Scalable Document Layout Annotation Pipeline for Layout Detection
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (37) · 30907-30915
DOI: 10.1609/aaai.v40i37.40349 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Feijiang Han und 7 weiteren Autor:innen.
Abstract (Originalsprache)
General-purpose Vision-Language Models (VLMs) are increasingly integral to modern AI systems for document understanding, yet their ability to perform fine-grained layout analysis remains severely underdeveloped. Overcoming this limitation requires large-scale, high-fidelity training datasets. However, current annotation methods that rely on parsing rendered PDFs are costly, error-prone, and difficult to scale. We propose a different paradigm: extracting ground-truth layout directly from the LaTeX compilation process rather than the final PDF. We present LaTeX2Layout, a generalizable procedural pipeline that recovers pixel-accurate bounding boxes and reading order from compiler traces. This enables the generation of a 140K-page dataset, including 120K programmatically generated synthetic variants that more than double the layout diversity of real-world data. Using this dataset, we fine-tune an efficient 3B-parameter VLM with an easy-to-hard curriculum that accelerates convergence. Our model achieves Kendall's tau=0.95 for reading order and mAP@50=0.91 for element grounding, delivering nearly 200% relative improvement over strong zero-shot baselines such as GPT-4o and Claude-3.7.
Zitieren
APA
Han, F., Wang, Z., Wang, B., Liu, X., Cheung, S., Rao, D., Callison-Burch, C., & Ungar, L. (2026). LaTeX2Layout: High-Fidelity, Scalable Document Layout Annotation Pipeline for Layout Detection. Proceedings of the AAAI Conference on Artificial Intelligence, 40(37), 30907-30915. https://doi.org/10.1609/aaai.v40i37.40349
Vancouver
Han F, Wang Z, Wang B, Liu X, Cheung S, Rao D, et al. LaTeX2Layout: High-Fidelity, Scalable Document Layout Annotation Pipeline for Layout Detection. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(37):30907-30915. doi:10.1609/aaai.v40i37.40349
BibTeX
@article{Han2026,
author = {Han, Feijiang and Wang, Zelong and Wang, Bowen and Liu, Xinxin and Cheung, Skyler and Rao, Delip and Callison-Burch, Chris and Ungar, Lyle},
title = {LaTeX2Layout: High-Fidelity, Scalable Document Layout Annotation Pipeline for Layout Detection},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {37},
pages = {30907--30915},
doi = {10.1609/aaai.v40i37.40349}
}
EUVIMED