Professors Isabel Barbancho and Lorenzo J. Tardón from University of Malaga visit MIB and give talks.

09.11.2017 | Hella Kastbjerg

ISABEL BARBANCHO

TITLE

Music Information Retrieval & some of its applications

ABSTRACT

Music Information Retrieval (MIR) is interdisciplinary science that deals with processing, analysing, searching, organizing and accessing music-related data. Music becomes music after being processed by the human mind and each person perceives the music in a different and complex way. Therefore, MIR embraces the complexity and diversity of music by pursuing research in ideas and applications that aim to enhance the way in which we interact with music.

MIR is a truly multidisciplinary area, involving researchers, developers, educators, librarians, students and professionals from the disciplines of musicology, cognitive science, library and information science, computer science, electrical engineering and many others.

In this talk, an overview of different challenges identified to be solved by MIR will be presented. Then, some ideas of how to perform music processing will be given. Finally, some of the development already done by the ATIC Research Group from the University of Málaga (Spain) will be presented.  

LORENZO J. TARDON

TITLE

Audio Signal: Characterization and Classification

ABSTRACT

Automatic audio and signal analysis relies on the identification and selection of features carrying certain semantic information. The specific definition of the features to employ is of main importance for the behaviour of later stages. Additional feature selection strategies arise to facilitate the good performance of the whole system; these are aimed at removing outliers, discarding features of reducing dimensionality. Finally, the actual interpretation and classification of signals can be carried out by means of different techniques based on diverse foundations.

In this talk, a brief overview of signal features focused on audio and music signals will be given, together with some simple ideas and techniques commonly used for feature selection and classification.