Using Handwritten Text Recognition: An Introduction to Transkribus (A Workshop)
Sponsoring Organization(s)
Recognition and Enrichment of Archival Documents (READ)
Organizer Name
Tobias Hodel
Organizer Affiliation
Staatsarchiv des Kantons Zürich
Presider Name
Louise Seaward
Presider Affiliation
Univ. College London
Paper Title 1
READ: An Introduction
Presenter 1 Name
Louise Seaward
Paper Title 2
Preparing Documents in Transkribus
Presenter 2 Name
Maria Kallio
Presenter 2 Affiliation
Kansallisarkisto
Paper Title 3
Training HTR Models and Searching in Large Collections
Presenter 3 Name
Tobias Hodel
Start Date
12-5-2019 10:30 AM
Session Location
Schneider 1235
Description
Algorithms deciphering and searching handwritten documents will broaden our access to and our insight into medieval sources tremendously. Due to the development of machine learning (based on neural networks) the last few years have seen an incredible improvement in the recognition of layout and handwriting, so that the technology can now be used by scholars in the humanities for their everyday work. Although training data is still required to recognize text, the use of the technology is becoming more and more worthwhile, especially for medievalists and early modernists, as better and better models are being developed. The use of Transkribus is already suitable not only for large projects that work with hundreds of manuscript pages, but also for small-scale applications such as dissertations and master theses.
Transkribus (https://transkribus.eu/Transkribus/) is an open source platform for the automated recognition, transcription and searching of handwritten historical documents. Transkribus is part of the EUfunded Recognition and Enrichment of Archival Documents (READ) (http://read.transkribus.eu/) project. The core mission of the READ project is to make handwritten material more accessible through the development and dissemination of Handwritten Text Recognition (HTR) and other cutting-edge technologies.
The Transkribus workshop is aimed at researchers and students who are interested in the transcription, searching and publishing of historical documents. It will introduce participants to the technology behind the READ project and demonstrate the Transkribus transcription platform.
Participants are asked to create a Transkribus account and install Transkribus on their laptops in advance of the workshop. They should bring their laptops along to the workshop. Wi-Fi is necessary. Tobias Hodel
Using Handwritten Text Recognition: An Introduction to Transkribus (A Workshop)
Schneider 1235
Algorithms deciphering and searching handwritten documents will broaden our access to and our insight into medieval sources tremendously. Due to the development of machine learning (based on neural networks) the last few years have seen an incredible improvement in the recognition of layout and handwriting, so that the technology can now be used by scholars in the humanities for their everyday work. Although training data is still required to recognize text, the use of the technology is becoming more and more worthwhile, especially for medievalists and early modernists, as better and better models are being developed. The use of Transkribus is already suitable not only for large projects that work with hundreds of manuscript pages, but also for small-scale applications such as dissertations and master theses.
Transkribus (https://transkribus.eu/Transkribus/) is an open source platform for the automated recognition, transcription and searching of handwritten historical documents. Transkribus is part of the EUfunded Recognition and Enrichment of Archival Documents (READ) (http://read.transkribus.eu/) project. The core mission of the READ project is to make handwritten material more accessible through the development and dissemination of Handwritten Text Recognition (HTR) and other cutting-edge technologies.
The Transkribus workshop is aimed at researchers and students who are interested in the transcription, searching and publishing of historical documents. It will introduce participants to the technology behind the READ project and demonstrate the Transkribus transcription platform.
Participants are asked to create a Transkribus account and install Transkribus on their laptops in advance of the workshop. They should bring their laptops along to the workshop. Wi-Fi is necessary. Tobias Hodel