Image Systems and Urban Spatio-Temporal Navigation

Over the past decade, significant progress has been made in developing semi-automatic information extraction pipelines for large collections of cultural heritage digital images. This presentation will center around Image Systems, a novel formalism that allows for the conversion of digital image collections into structured datasets that are optimised for computation. An Image System comprises a graph of interconnected digital images, guided by a set of rules determining when two images ought to be linked. Unlike conventional approaches based on metadata, Image Systems prioritize the nature of the relationship between images, and the types of computations for extracting relevant information. The presentation will showcase fruitful results of this approach in the areas of urban spatiotemporal navigation and automated discovery techniques for the digital humanities.

Frédéric Kaplan, professor at the École Polytechnique Fédérale de Lausanne (EPFL), directs the Institute of Digital Humanities and the College of Humanities, which brings together research and teaching in the humanities and social sciences at EPFL. He is also president of the Time Machine Organization, a non-profit entity of more than 600 institutions. His work, at the crossroads of information science and humanities, has led to more than a hundred scientific publications, exhibitions in several major museums (the Venice Architecture Biennale, the Grand Palais, the Centre Pompidou in Paris and the Museum of Modern Art in New York) and a dozen books, translated into several languages.

Please find the video registration of the event on our VIMEO CHANNEL: https://vimeo.com/manage/videos/802628829

The lecture is part of the Winter School: Datathink 2023: The City as Collection

Scientific Organization: Dario Negueruela del Castillo