Milanka Matić-Chalkitis (MultiHTR project) · PyLaia · Published May 8, 2023

Stolze-Schrey_combined

Text Recognition

Description

This is the first model version for the German Stolze-Schrey shorthand system, which is largely based on synthetic, but also natural training data. Synthetic training data, electronically available longhand texts converted into the corresponding shorthand system, were taken from the following page:https://www.vsteno.ch/php/library.php. The natural training data consists of a few pages of private field post in letter and postcard format from 1939. The model was trained by Milanka Matić-Chalkitis as part of the MultiHTR project (project leader: Prof. Dr. Achim Rabus) at the Department of Slavic Languages and Literatures of the University of Freiburg (Germany). It offers initial assistance for handwritten documents written in Stolze-Schrey shorthand by automatically generating a kind of 'pre-transcription'.

Try this model

Use this modelOpen in Transkribus
Low error rate9.5% CER

Character Error Rate (CER) measures the percentage of characters incorrectly recognised. Lower is better. This model scored 9.5% on its validation set. As a rule of thumb, a CER below 10% is considered good for most handwritten material.

Measured on the model's own validation data. Results on your documents may differ depending on handwriting style, document condition, language, and how closely your material resembles the training data.

Words23,023
Lines2,292
Training Pages167
Model ID52058