REPUBLIC 7 – 17th and 18th century Dutch

Free Public AI Model for Handwritten Text Recognition with Transkribus

REPUBLIC 7 – 17th and 18th century Dutch

This model is based on the handwritten resolutions of the Dutch States-General (1576-1796). It has been created during the project ‘REPUBLIC’, which aims to provide an online searchable edition of the Reolutions.
For more information regarding the background of the model and how to cite it, please have a look here.

During the first phase, Huygens ING have created a ground truth corpus of around 1,000 pages, with which several HTR-models have been trained. The latest model, which Huygens ING proudly releases for common use, is performing at a staggering 2.99% Character Error Rate. With the help of volunteers Huygens ING aims at lowering this percentage even further during the next phase of the project. Because of the excellent HTR results they are confident that the ultimate goal, making this important source (re)searchable from all sorts of perspectives, will become reality.

A Resolution from 1598 © Project REPUBLIC

Model Overview

Name:
Republic_7
Creator:
REPUBLIC Project (Huygens ING)
Model ID:
23967
Century:
17th, 18th
Languages:
Dutch
Script:
Latin alphabet
Engine:
HTR+
Material:
Handwritten
CER on validation set:
2.99 %

Republic_7 is freely available to everyone

Get started with Transkribus and use it for your own Material
You can use this model to automatically transcribe Handwritten documents with Handwritten Text Recgnition in Transkribus. This model can be used in the Transkribus Expert Client as well as in Transkribus Lite.
This AI model was trained to automatically convert text from images of historical Latin alphabet documents into editable and searchable text.