Benutzerspezifische Werkzeuge

With Leonardo Impett, Bibliotheca Hertziana – Max Planck Institute, Rome and Ralph Ewerth, University of Marburg | Hosted by Johannes Griebel and Georg Schelbert

The ongoing kunst.bild.daten project, which aims to digitally catalogue the ZI’s photo library, builds on existing data in the form of handwritten and typewritten inventories. To manage exceptionally large and heterogeneous collections, LLMs are used – partly in complex prompt cascades – to generate digital representations of the inventories and, furthermore, to disambiguate person and object data and link it to authority files.

What has so far been excluded from these processes is working directly on the images. Two experts in computer vision, Leo Impett (currently research group leader for the project ‘Machine Visual Culture: AI and the Stories of Seeing’ at the Bibliotheca Hertziana, Rome) and Ralph Ewerth (Professor specialising in multimodal modelling and machine learning at the University of Marburg) will give insights in their work and discuss with us the potential prospects for the use of computer vision in our photo library.

// Leonardo Impett is Research Group Leader at the Bibliotheca Hertziana, Associate  Professor in Digital Humanities at the University of Cambridge,  Bye-Fellow of Selwyn College, and Research  Fellow of the British School at Rome. He has a background in  information engineering and machine learning for the humanities, and was  previously Assistant Professor of Computer Science at Durham  University. His research explores the intersections, in both  directions, between artificial intelligence and art history or visual  studies. He has been working both on computer vision for art history,  and on AI's "ways of seeing", since 2015.  

// Ralph Ewerth studied computer science, with a minor in psychology, in Frankfurt and Marburg, and obtained his PhD in Marburg in 2008 with a thesis on machine learning for video analysis. From 2012 to 2015, he was a professor of image processing and media technology at Jena University of Applied Sciences, before he became a professor at Leibniz Universität Hannover in a joint appointment with TIB – Leibniz Information Centre for Science and Technology, where he heads the Visual Analytics research group. Since 2025, Ralph Ewerth has been a professor of multimodal modeling and machine learning at Philipps-Universität Marburg and is a member of the Hessian Center for Artificial Intelligence (hessian.AI).

---

PARTICIPATION: The event will be held in person only and is open to project members, ZI staff, and interested guests. Please register at kunst.bild.daten@zikg.eu. The event will be conducted in English. // TEILNAHME: Die Veranstaltung findet nur vor Ort statt und richtet sich an Projektmitglieder, Mitarbeitende des ZI und interessierte Gäste. Anmeldung bitte unter  kunst.bild.daten@zikg.eu. Die Veranstaltung findet auf englischer Sprache statt