Centre of Excellence for Automatic Text Recognition (ATR)

Through its Centre of Excellence for Automatic Text Recognition (ATR), the ULB Tirol provides a point of contact for advice, support and networking in the field of digital humanities. We support researchers, projects and students at the University of Innsbruck throughout the ATR workflow – from the initial idea through to test runs and quality assurance, right up to the delivery of research outputs.

Projects are not carried out independently by ULB Tirol; instead, we provide advice and support on an ad hoc basis – where possible, even before the project begins and a project proposal is submitted, and in any case following consultation at defined stages of the project. To this end, we are happy to make our expertise and internally used tools available, but we also work closely with the Digital Humanities Research Centre, the ZID Service CentreDigital Research Servicesand the respective data stewards.

Are you interested or do you have a specific enquiry?
Please feel free to contact us with a brief description of your project, the source material and the support you require.

Contact: Mirjam El Attal | Digital Services Department | ulb-atr@uibk.ac.at

Infografik ATR

Infografik zu Services der ATR-Kompetenzstelle

Infografik zu den Services der Kompetenzstelle für Automatische Texterkennung. 
1) Beratung zu Projektbeginn: Bedarfsanalyse, Aufwandsschätzung und Empfehlungen zu wichtigen Entscheidungspunkten im gesamten ATR-Prozess.
2) Evaluation & Qualitätsbewertung: Durchführen von Testläufen, Stichprobenbewertungen, Review von Zwischenergebnissen sowie Beratung zur Erstellung von Ground-Truth-Material (Umfang, Qualität, Dokumentation).
3) Toolauswahl & Dokumentation: Orientierung zu ATR-Tools und Best Practices, Bereitstellung ausgewählter interner Services sowie Verweise auf nützliche externe Anlaufstellen; Unterstützung bei Auswahl von Formaten, Parametern und Workflow-Dokumentation.
4) Workshops & Schulungen: Einführungsworkshops, thematische Vertiefungen – z.B. zum Modelltraining in Transkribus – und toolspezifische Schulungen nach Bedarf.
5) Daten verfügbar machen: Beratung zur Einbindung von Forschungsoutput in bestehende Bibliotheksservices und zu möglichst nachhaltigen Workflows für Langzeitverfügbarkeit.
6) OCRmyPDF: Browserbasierter Webservice für alle Universitätsangehörigen, um selbstgescannte Dateien durchsuchbar zu machen. Link zum OCR-Webservice: ulb-ocr.uibk.ac.at

OCRmyPDF – how it works

  • What is it?The service automatically makes self-scanned files searchable using OCR.

  • Where to use it?ulb-ocr.uibk.ac.at Log in with your UIBK account and two-factor authentication. 

  • What to upload?PDF scans that are as legible as possible (straight, not too dark or too light).

  • Select language: If your document contains multiple languages, add all the required languages in the ‘Language for job’ field – a maximum of 5 languages can be selected.

  • Old-style typeface (Fraktur): Tick the‘Show all languages’option and select a Fraktur font, e.g. German – Fraktur.

  • Start & Duration: Start the job by uploading the file; depending on the file size, this may take a little time – maximum file size is 2GB per job. 

  • Results: Download as a single PDF or Word file (.docx). Alternatively, you can download a ZIP file containing all available output formats: DOCX, PDF, TXT, XML.

Important: Uploaded files are automatically deleted from the server after 14 days. Please save the output files immediately.
Help: ulb-atr@uibk.ac.at

Nach oben scrollen