The primary, overall objective of the HUGIN-MUNIN project is to develop new deep neural network technology with general, robust models that will enable large-scale use of HTR as part of the standard digitization process of Norwegian LAM collections without the need for massive and on- going manual annotation and model training.
In order to attain the primary project objective, new methods will be developed that go beyond traditional supervised machine learning, by using active learning, unsupervised learning, transfer learning, and zero-shot learning.
Key to the project will be an interdisciplinary collaboration between specialized research communities within document analysis and HTR, both nationally and internationally. This will expand Norwegian experience and competence in AI/autonomous systems expertise and enhance the innovative potential of the Norwegian LAM sector through transfer of knowledge and best practices.
Funded by the Norwegian Research Council within the IKTPLUSS initiative
December 22, 2023
Release of PyLaia v1.1.0 with a full public documentation PyLaia
October 18, 2023
Release of PyLaia v1.0.7 with better confidence computation with language model PyLaia
September 12, 2023
Release of PyLaia v1.0.6 with temperature scaling PyLaia
December 12, 2022
Release of PyLaia v1.0.3 with language model PyLaia
December 7, 2022
Confidence scores added to PyLaia
November 10, 2022
Release of Doc-UFCN line detection models on HuggingFace
May 25, 2022
Oral presentation a paper at DAS2022
May 20, 2022
Publication of our NorHand dataset on Zenodo
April 27, 2022
Kick-off meeting in Oslo