Navigating Multigraphic Written Artefacts

Project Group

This Project Group advances UWA’s commitment to digital innovation by utilizing AI and computer vision to enhance the study of multigraphic WAs, which include multiple sign systems like scripts and pictorial elements. Focusing on visual organization, the group aims to improve accessibility and analysis of large digitized cultural collections. Three case studies will demonstrate AI’s potential: analyzing French medieval Bible moralisée manuscripts, classifying subject-based Islamic manuscripts from West African digital archives, and navigating a vast archive of German-Jewish images at the National Library of Israel. These studies will explore how visual elements are organized, map large corpora, and facilitate access to uncatalogued collections. Researchers will identify unique patterns for each study and analyze their visual, spatial and semantic relationships across each collection. Outcomes include novel AI approaches evaluated on vision-language datasets derived from the aforementioned studies, an interactive visual interface for engaging with the developed approaches, and several articles on the developed methods, created datasets, and explored case studies. This collaboration between humanities and computer science aims to establish a new model for exploring intricate manuscript collections.