PereStruct YOLO Dataset

Parsing historical documents with complex, non-standard layouts remains a fundamental bottleneck in large-scale archival digitization. Unlike modern typography, historical newspapers exhibit severe physical degradation and highly irregular page structures that confound even state-of-the-art vision-language models, presenting severe out-of-distribution challenges. We address this gap with an automated pipeline specifically designed for parsing historical newspapers, documents characterized by particularly intricate multi-column layouts. Our approach combines a fine-tuned YOLO architecture for layout analysis and block detection, trained on 1,426 fully human-annotated scanned pages, with a novel semantic assembly module that reconstructs articles by jointly modeling lexical-semantic similarity via TF-IDF, visual embeddings from our fine-tuned YOLO, and geometric layout constraints. This multi-modal integration yields state-of-the-art performance, achieving an F1 score of 0.904 on block-to-article mapping. Notably, end-to-end evaluation against vision-language models (Qwen3.6-35B-A3B and Qwen3.6-Plus) demonstrates that PereStruct achieves substantially higher fidelity (BLEU approximately 0.96 vs 0.34), validating that modular architectures excel where generic VLMs fail on complex historical layouts. To support reproducibility and advance research in this domain, we release both the training corpus of 599 annotated pages and a curated PereStruct benchmark of 93 pages with expert-verified ground-truth block-to-article mappings. This framework establishes a robust foundation for high-fidelity digitization and semantic reconstruction of complex archival materials.

We curate a large-scale training dataset comprising 1,426 high-resolution scanned pages from the N.A. Nekrasov Library

collection, annotated for layout detection in standard YOLO format. These annotations were produced in collaboration

with the Yandex Crowd team, utilizing a rigorous multi-stage quality control protocol to ensure consistent bounding

box placement across the heterogeneous visual styles of the era. For each page, the dataset provides normalized

bounding box coordinates (xcenter,ycenter,width,height) alongside categorical labels spanning four classes: Title, Plain

Text, Figure, and Figure Caption.

While our layout detection model was trained on the full corpus of 1,426 images to capture the full spectrum of

observed complexity, due to copyright restrictions, we publicly release a subset comprising 599 annotated pages. This

corpus provides a robust foundation for training domain-specific detectors and evaluating performance on irregular

multi-column flows and varying degradation patterns.