Transkribus Webinar #2 - Advanced Use by Annemieke Romein

Creating transcriptions can be a tedious job. The computer tool Transkribus can ease the process after some training. In this digital workshop, we will commence with the knowledge you gained in the basic training. You will now learn about judging the quality of a model, structure tags and automatic training of structure through P2PaLA, combining existing transcriptions with images (T2I), future options with tables and NERs-tagging.
This workshop was organised through a Zoom-meeting. The workshop is in English and is provided in by Annemieke Romein, a postdoctoral researcher at Huygens ING and experienced Transkribus-user.

The slides can be found at Zenodo: https://doi.org/10.5281/zenodo.3834205

Intro to the webinar 00:00:00-00:01:53
Models intro 00:01:54-00:02:11
- Reading the graph 00:02:12-00:09:27
How to compare the quality of models for your documents? 00:09:28-00:28:23
- Compare button 00:09:28-00:16:09
- Compare text-versions 00:16:10-00:18:53
- Compare samples 00:18:54-00:21:27
- How to pick a model? 00:21:28-00:24:42
- Questions 00:24:43-00:28:23
Text2Image 00:28:25-00:33:25
- Methods 1 & 2 (with/out manual LA) 00:33-26-00:47:45
- More manual (complete control) 00:47:46-00:52:34
- Questions and additional showing of method 1 00:52:34-01:06:05
Tags 01:06:06-01:06:39
- How to add (document) metadata? 01:06:40-01:08:50
- How to tag (NERs) in Transkribus manually) 01:08:51-01:12:20
- How to do structure-tagging 01:12:21-01:18:20
P2PaLA 01:18:21-01:36:22
Tables (what will the future/ NaverLABS bring). 01:36:23-01:47:11