Multimedia Information Retrieval and Music Processing (Former Research Group)

Vision and Research Strategy

Generally speaking, our vision is to provide novel technologies that allow for organizing, understanding, and searching multimodal information in a robust, efficient and intelligent manner. Based on a rigorous mathematical foundation, the general research mission of our group is to develop generic methods that facilitate efficient and deformation-tolerant retrieval of object constellations in large multimedia datasets. Important aspects concern the design of suitable features, the notion of similarity used to compare multimedia objects, as well as data organization.

Our group attaches great importance to dealing with research problems of practical relevance. To demonstrate the potential of the developed analysis and retrieval techniques, prototypical implementations of user interfaces and search engines have been and will be developed. Working with real-world data, many cross connections to various other research fields have been established.

In the motion domain, we are collaborating with experts from computer graphics, computer vision, and sport sciences. The music domain is interdisciplinary par excellence. Here, our group is cooperating with scientists from library science and musicology. Recently, a long-term collaboration with the University of Music, Saarland, has been established to investigate the potential of computer-based methods for music education.

Composition of Group

The research group Multimedia Information Retrieval and Music Processing headed by Meinard Müller joined the Cluster of Excellence in December 2007. Its general research mission is to advance the development of techniques and tools for analyzing, structuring, retrieving, navigating, and presenting time-dependent data streams with a focus on music data and motion data. Dealing with two different multimedia domains, the group's research falls into two major areas:

  1. Music information retrieval, digital signal processing, and music processing.
  2. Content-based retrieval and analysis of three-dimensional human motion capture data.

During 2008, four doctoral students from different disciplines have joined the research group: Dipl.-Math. Verena Konz and Dipl.-Ing. Peter Grosche (both funded by the Cluster of Excellence) are working in the music domain, while Dipl.-Inf. Andreas Baak and Dipl.-Inf. Thomas Helten (both funded by individual DFG grants) are working in the motion domain.

Research Topics and Achievements

Prizes and Awards

Prof. Dr. Meinard Müller

Prof. Dr. Meinard Müller

Meinard Müller is a full professor at the International Audio Laboratories Erlangen (AudioLabs), a joint institution of Fraunhofer IIS and University of Erlangen-Nürnberg.

Fon: +49 9131 85 20504