Winter School: HTR of Historical Sources, 2026
→
Europe/Vienna
Virtual
Virtual
Description
The use of AI for reading historical sources (known as Automated or Handwritten Text Recognition) is becoming an expected and necessary part of many DH projects. Nevertheless, there are still many challenges, from choosing between a growing number of options for HTR, to learning how to prepare ground truth and train models, and to navigating new methodological issues.

The Winter School: HTR of Historical Sources, 2026 (further as HTR Winter School) started in 2022 and has evolved along with technological developments and the current state of research. Most importantly: Its participants learn not only the basics of HTR, but contribute to growing datasets for particular languages and scripts.
Working groups
This year, there will be working groups for the following languages/scripts:
- Carolingian Latin
- Late Medieval Latin
- Early Modern German
- Byzantine Greek
- Syriac
- Church Slavonic
- Old Icelandic
Each of these groups will work on its own materials and has its own requirements for participation. Please check them before registering.
Only participants who fulfill their tasks in the language/script groups will receive the final digital certificate.
Timetable
This year, the Winter School will take place online only (with the exception of the Old Icelandic group that will have an additional in-person meeting), resuming in-person participation in 2027.
There are five virtual sessions, always on Friday from 2PM to 5PM, Central European Time:
- 2nd October
- 16th October
- 30th October
- 13th November
- 27th November
Each session consists of a plenary part (lecture) and a working group part. For all sessions except the last one, the plenary part will be 45 minutes long and the group part 2 hours long.
The last session will be dedicated to presenting the results.
Call for applications
Call for application is open until 11th September 2026.
You will be notified on 18th September, whether you have been accepted.
Registration
Call for application