Devanagari mixed M1A

Free Public AI Model for Handwritten Text Recognition with Transkribus

Devanagari mixed M1A

This HTR+ model was trained for various Devanagari print types based on approximately 26,000 words. The training material is comprised of letterpress printings from the Naval Kishore Press (Lakhnau, North India) from the late 19th and early 20th century in the Hindi, Sanskrit, Braj Bhasha and Awadhi languages. The model was created as part of the project “Naval Kishore Press – digital”, which was financially supported by the German Research Foundation (DFG) until 2021.

Dieses HTR+ Modell wurde für verschiedene Devanagari-Drucktypen auf der Grundlage von ca. 26.000 Wörtern trainiert. Beim Trainingsmaterial handelt es sich um Buchdrucke der Naval Kishore Press (Lakhnau, Nordindien) aus dem späten 19. und frühen 20. Jahrhundert in den Sprachen Hindi, Sanskrit, Braj Bhasha und Awadhi. Das Modell wurde im Rahmen des Projekts “Naval Kishore Press – digital” erstellt, das von der Deutschen Forschungsgemeinschaft (DFG) bis 2021 finanziell unterstützt wurde.
https://digi.ub.uni-heidelberg.de/sammlungen/suedasien/navalkishore.html

Model Overview

Name:
Devanagari mixed M1A
Creator:
Nicole Merkel-Hilf (UB Heidelberg)
Model ID:
45909
Century:
19th, 20th
Languages:
Hindi, Braj Bhasha, Sanskrit, Awadhi
Script:
Devanagari
Engine:
PyLaia
Material:
Print
CER on validation set:
2.20 %
Simply upload a picture and test this model

By uploading an image, you accept our terms and conditions and our privacy policy

Devanagari mixed M1A is freely available to everyone

Get started with Transkribus and use it for your own Material
You can use this model to automatically transcribe Print 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 Devanagari documents into editable and searchable text.