iART

The project iART develops methods for the realization of an interactive image retrieval system, which allows users to search and explore large image databases of art history.

Description

The goal of the iART project is to develop an open Web platform, which is able to index large amounts of image data from art history and related disciplines and make them searchable. Specific methods will be designed for content-based search that take into account color, texture, and shape as well as the semantic content of the image. In order to achieve this goal, a generic lexicon of visual concepts will be created together with the project partners and the technical feasibility of the concepts will be evaluated with regard to the various object types. In addition, the system will include functions to cluster and structure large sets of images based on various definitions of similarity.

Cooperation

• Paderborn University, Department of Computer Science, Intelligent Systems and Machine Learning Group, Prof. Eyke Hüllermeier
• Ludwig-Maximilians-Universität München, Lehrstuhl für Mittlere und Neuere Kunstgeschichte, Prof. Hubertus Kohle