In This Section
- Graham Blyth Celebrates 20 years of Organ Concerts at AES 135
- Audio and Video Downloads Now Available
- Berlin Convention Registration Now Open!
- Mobile App is Also Ready to Download
- Call for Nominations Deadline Approaching
- Deadline is February 17th
- 135th AES Convention Hits A Five-Year High
- Convention takes a bite out of the Big Apple and reminds the industry that “If It’s About Audio, It’s At AES”
Music Structure Analysis
International Audio Laboratories Erlangen, Germany
One of the attributes distinguishing music from other sound sources is the hierarchical structure in which music is organized. Individual sound events corresponding to individual notes form larger structures such as motives, phrases, and chords, and these elements again form larger constructs that determine the overall layout of the composition. One important goal of music structure analysis is to divide a given audio recording into temporal segments that correspond to musical parts and to group these segments into musically meaningful categories. One challenge is that there are many different criteria for segmenting and structuring music. This results in conceptually different approaches, which may be loosely categorized in repetition-based, novelty-based, and homogeneity-based approaches. Furthermore, one has to account for different musical dimensions such as melody, harmony, rhythm, and timbre. In this talk, I will give an overview of current approaches for the computational analysis of the structure of music recordings, which has been a very active research problem within the area of music information retrieval. As one example, I present a novel audio thumbnailing procedure to determine the audio segment that best represents a given music recording. Furthermore, I show how path and block structures of self-similarity matrices, the most important tool used in automated structure analysis, can be enhanced and transformed. Finally, I report on a recent novelty-based segmentation approach that combines homogeneity and repetition principles in a single representation referred to as structure feature.
Bio of presenter
Meinard Müller studied mathematics (Diplom) and computer science (Ph.D.) at the University of Bonn, Germany. In 2002/2003, he conducted postdoctoral research in combinatorics at the Mathematical Department of Keio University, Japan. In 2007, he finished his Habilitation at Bonn University in the field of multimedia retrieval writing a book titled "Information Retrieval for Music and Motion," which appeared as Springer monograph. From 2007 to 2012, he was a member of the Saarland University and the Max-Planck Institut für Informatik leading the research group "Multimedia Information Retrieval and Music Processing" within the Cluster of Excellence on Multimodal Computing and Interaction. Since September 2012, Meinard Müller holds a professorship for "Semantic Audio Processing" at the International Audio Laboratories Erlangen, which is a joint institution of the Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) and the Fraunhofer-Institut für Integrierte Schaltungen IIS. His recent research interests include content-based multimedia retrieval, audio signal processing, music processing, music information retrieval, and motion processing.