Skip to main content

Automatic Text Recognition (ATR) - Step 5: Text Recognition and Post-ATR Correction

Authors
Topics:

This session breaks down the essentials of creating accurate models for Automatic Text Recognition (ATR), including understanding ground truth data. Perfect for enhancing your ATR skills, the tutorial equips you with the knowledge to improve text extraction from heritage materials.

Learning Outcomes

After completing this resource, learners will be able to:

  • Build and train models to recognise and interpret text within images.
  • Apply concepts of machine learning specific to ATR to improve recognition accuracy.
  • Create ground truth data for training and validating ATR models.
  • Evaluate the effectiveness of different ATR models based on their output quality.

You can read the blogpost (available in English, French, and German), and watch our video (with subtitles in English, French, and German) embedded in the post.

Interested in learning more?

Check out "Automatic Text Recognition - Step 5: Text Recognition and Post-ATR Correction

Go to this resource

Cite as

Floriane Chiffoleau and Sarah Ondraszek (2024). Automatic Text Recognition (ATR) - Step 5: Text Recognition and Post-ATR Correction. Version 1.0.0. Edited by Anne Baillot and Mareike König. Deutsches Historisches Institut Paris. [Training module]. https://harmoniseatr.hypotheses.org/226

Reuse conditions

Resources hosted on DARIAH-Campus are subjects to the DARIAH-Campus Training Materials Reuse Charter

Full metadata

Title:
Automatic Text Recognition (ATR) - Step 5: Text Recognition and Post-ATR Correction
Authors:
Floriane Chiffoleau, Sarah Ondraszek
Domain:
Social Sciences and Humanities
Language:
en
Published:
5/10/2024
Content type:
Training module
Licence:
CCBY 4.0
Sources:
DARIAH
Topics:
Editing tools, Machine Learning, Automatic Text Recognition
Version:
1.0.0