On October 23, the RegInfra team delivered a workshop on Semantic Image Annotation in cooperation with LECTIO and Dr. Rainer Simon, developer of IMMARKUS.
In this workshop, participants gained conceptual and practical insight into how digital image annotation can be integrated into their research. We mainly worked with IMMARKUS, a semantic image annotation service designed by the presenters that allows users:
- to assemble and organize images from their own image collections or streamed as high-quality zoomable images from libraries, museums, and archives;
- to create custom ontologies to identify and annotate specific image regions;
- to add relationships between elements in and across the images.
IMMARKUS supports commercial, open, and custom-designed AI models to select image parts or transcribe and translate text in various languages. From the built-in knowledge graph, users can visualize and query image annotations, relationships, and metadata, and export results for further data analysis.
This workshop consists of three parts:
- Conceptual introduction – an overview of annotation as a research method to analyze a wide range of primary sources, including manuscript and print text and graphics, maps, paintings, photographs, archeological diagrams, and all types of digitized objects.
- Hands-on practice – a guided demonstration of IMMARKUS, covering the creation of image corpora, annotation with various AI models, data model design, data visualization and export.
- Moderated discussion – a conversation with Dr. Rainer Simon about developing IMMARKUS and collaborative image annotation in liiive.
More information and registration: https://www.kuleuven.be/lectio/events/semantic-image-annotation