historical-documents

Historic(al) Maps: Meta guidance, tools, repositories, databases, search engines, and online resources for the exploration of Historic and Historical Maps.

A Mongoose plugin that archives document diffs and manages document history.

Data for the HIPE 2022 shared task.

Benchmarking experiments on LLM-based post-correction of historical OCR, HTR, and ASR transcripts.

OCR-D compliant toolset for optical layout recognition on historical german-language documents published in Brazil

🔬 Impresso Datalab Notebooks

Official PyTorch implementation of "Investigating the Effect of using Synthetic and Semi-synthetic Images for Historical Document Font Classification" - DAS 2021

Implementation of PHOSNet and Pho(SC)Net for Word Recognition in Historical Documents. Implemented using Tensorflow 2.x

Auditable page-level run ledger for OCR and document extraction

Thulium is a production-ready Python library for offline handwritten text recognition (HTR) supporting 52+ languages across Latin, Cyrillic, Greek, Arabic, Hebrew, Devanagari, Chinese, Japanese, Korean, and Georgian scripts.

A python-based pipeline to query the USGS Historical Topographic Map Collection (HTMC) and to download, spatially aggregate, and mosaic large amounts of individual map sheets.

Hosting for various texts to be used in NLP projects (et al)

🚀 The frontend application of the Impresso WebApp

The code for the paper "Historical Printed Ornaments: Dataset and Tasks" (ICDAR 2024).

Tai Le historical document image binarization dataset, TLHDIBD2021

Data for the HIPE 2026 shared task (CLEF 2026 Evaluation Lab)

This project offers an advanced Optical Character Recognition (OCR) solution specialized for Chinese historical documents, uniquely addressing complex layouts and reading orders.

Post-processing filter for (Named) Entity Linking

Semi-automated digitization of historical handwritten tabular records (JCDL'24). OCR + computer vision pipeline for archival beekeeping forms.

GPU-enhanced OCR pipeline for historical newspapers using Google Vision or Tesseract