Feature-rich text transcription editor for document images

Quickly produce a sufficient amount of ground-truth data for training your own AI model. The quickest way to do this is to recognise your material with a public model and then correct the results, instead of typing out everything by hand. By repeating this process a number of times and adding more training data to your model, you can train increasingly accurate AI models to recognise your material.

Within Transkribus, you can easily transcribe your documents with the image side by side with the text editor. The automatic Layout Analysis creates a correspondence between the lines in the image and the lines in the text editor, so while you are transcribing, the image is automatically centred on the line you are working on.

Transcribe, edit, and annotate your documents with ease, and produce high-quality scholarly editions. These can be shared conveniently with your target audience by setting up a Transkribus Sites website – no need to export your data and re-import them into a separate solution, or to build your own website etc.