Published May 12, 2025
| Version v0.2
Model
Open
OICEN-HTR: Old Icelandic / Norse Handwritten Text Recognition
Authors/Creators
Description
This repository contains HTR models fine-tuned with CATMuS Medieval 1.6.0 on Old Icelandic manuscripts. It also contains the ground truth data, but not the images, as they are usually available on Handrit.is.
Currently it includes the ground truth data for Alexanders saga (AM 519 4 to), Codex Wormianus (AM 242 fol), parts of Möðruvallabók (AM 132 fol), as well as the separate and combined HTR models fine-tuned on these data.
These models are better suited for Old Icelandic manuscripts than CATMuS. They produce from moderate to good results.
Information on the combined HTR model (v0.1):
- Evaluation:
- Characters: 1081450
- Errors: 27999
- Character Accuracy: 97.41%
- Word Accuracy: 91.42%
- Annotations are based on Menota AM 519 a 4to (v. 1.0.5) by de Leeuw van Weene, Menota AM 242 fol (v. 0.9.9) by Karl Gunnar Johansson, Menota AM 132 fol (v. 1.0) by de Leeuw van Weene.
- Images are from Handrit.
For more information see the GitHub repository NKCZ/OICEN-HTR.
Files
NKCZ/OICEN-HTR-v0.2.zip
Files
(67.4 MB)
| Name | Size | Download all |
|---|---|---|
|
md5:a00807f42bdc3772fb363cd421c43315
|
67.4 MB | Preview Download |
Additional details
Related works
- Is derived from
- Publication: https://clarino.uib.no/menota/text/menota/AM-519a-4to (URL)
- Publication: https://clarino.uib.no/menota/text/menota/AM-132-fol-Njals-saga (URL)
- Publication: https://clarino.uib.no/menota/text/menota/AM-242-fol (URL)
- Model: https://zenodo.org/records/15030337 (URL)
- Is supplement to
- Software: https://github.com/NKCZ/OICEN-HTR/tree/v0.2 (URL)
Software
- Repository URL
- https://github.com/NKCZ/OICEN-HTR