muzykantov-2005 · PyLaia · Published February 21, 2025

Troitskiy v1.0

Text Recognition

Description

A model trained on the Troitskiy Parimeinik (National Library of Russia, Saint-Petersburg, Ф.304/I №4., https://clck.ru/3Mpt7F; transcription provided by the Manuscript corpus, https://clck.ru/3Mpt8c, Kalashnikov Izhevsk State Technical University) by Maxim Muzykantov (KISTU) for automatic transcription of Slavic uncial manuscripts from 11th and 12th centuries.

Try this model

Use this modelOpen in Transkribus
Low error rate6.25% CER

Character Error Rate (CER) measures the percentage of characters incorrectly recognised. Lower is better. This model scored 6.25% 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.

Words48,230
Lines10,941
Training Pages128
Model ID294093
Languages
Church Slavic
Centuries
13th c.14th c.