Munch: Video Retrieval system for Dutch national video archive

Multimedia aNalysis for Cultural Heritage (MuNCH). MuNCH focuses on knowledge enrichment by means of automated analyses of digital images and video. With the advent of digital communication, we live in the exciting times of broad and narrow casting through the Internet, of passive and active viewers, of direct or delayed broadcast, and of digital pictures being delivered in the museum or at home. At the same time, the picture and television archives turn digital. In these demanding times, the archives are likely to be swamped with information requests unless they swiftly adapt to partially automatic annotation and digital retrieval. The aim of this project is to provide faster and more complete access to pictures in cultural archives through digital analysis. MuNCH is a project of the University of Amsterdam, the Vrije Universiteit Amsterdam, the Netherlands Institute for Sound and Vision and the association Digitaal Erfgoed Nederland. Projectleader is Arnold Smeulders. Intelligent Systems Lab Amsterdam on the project:

In the MuNCH project - part of the CATCH projects by NWO - we aim to deliver a working video retrieval system to the Beeld en Geluid national (video)archive. A first version is planned for at the end of 2006 making use of the engines for video analysis and interaction. In the MuNCH, we cooperate with Maarten de Rijke (UvA) on natural language processing, with Guus Schreiber (VU) on ontologies of video material, and of course with The Netherlands Institute for Sound and Vision, Annemieke de Jong providing the content (essence + metadata).

Participants of Intelligent Systems Lab Amsterdam are:

• Jan van Gemert
• Jan-Mark Geusebroek
• Arnold Smeulders (contact MuNCH)
• Cees Snoek (contact TRECVID)
• Marcel Worring (contact MediaMill)

Filed in: Imaging, Multimedia Retrieval, Search.