Online, 15 December 2021

Hubertus Kohle & Stefanie Schneider
Ludwig Maximilian University of Munich

Assessing Visual Similarity: from Crowdsourced Taggings to AI-based Recommender Systems

The ACDH-CH Lecture 7.4 in cooperation with DArtHist Austria focusses on art history. Art-historical recognition processes are motivated by descriptions of similarities: in Heinrich Wölfflin’s case, these are conducted through form analysis, in Aby Warburg’s from the perspective of the iconol-ogist researching cultural studies. Wölfflin visualized similarities and differences in his Kunstgeschicht-liche Grundbegriffe on a double page with comparative examples, categorized in five binary opposites. Warburg designed his picture atlas plates in groups with similarity-related connections.

This lecture focuses on two computationally driven projects for evaluating similarities between art works, both developed at the Ludwig Maximilian University of Munich in recent years: the “ARTigo” gaming ecosys-tem, which consists of different types of crowdsourcing games that aim to collect rich annotations, and “iART”, an AI-fueled search engine that directly addresses image data.

Hubertus Kohle

Hubertus Kohle is a professor of art history at the Ludwig Maximilian University of Munich with inter-ests in 18th to 20th century German and French art and digital art history. He published books on French 18th to 19th social art history, Adolph Menzel, overviews on German mid-19th century art, the computer in art history, and Denis Diderot’s art theory.

Stefanie Schneider

Stefanie Schneider is a research assistant at the Ludwig Maximilian University of Munich. Trained as a statistician and computer scientist, she has been working in the digital humanities since 2016. Her re-search topics include digital art history, cultural analytics, digital film studies with a focus on genre theory, and research software engineering.


15 December 2021 – 16:00


Online (Zoom) 


Florian Wiencek