Time: 2:00 - 3:00pm Venue: Eng. 2.09 Engineering Building, Queen Mary University of London, Mile End Road, London, E1 4NS
A fundamental problem with nearly all work in music genre recognition (MGR) is that evaluation lacks validity with respect to the principal goals of MGR. This problem also occurs in the evaluation of music emotion recognition (MER), not to mention autotagging. Standard approaches to evaluation, though easy to implement, do not reliably differentiate between recognizing genre or emotion from music, or by virtue of confounding factors in signals (e.g., equalization). I demonstrate many problems with MGR and MER system evaluation, and conclude with recommendations.
Bob L. Sturm is an associate professor in the Audio Analysis Lab (http://www.create.aau.dk/audio/) at Aalborg University. He is currently funded by a two-year Independent Postdoc Grant from the Danish Agency for Science, Technology and Innovation. His research interests are: digital signal processing and machine learning for audio and music signals, algorithms for sparse approximation and compressive sampling, and evaluation. Bob received the B.A. degree in physics from University of Colorado, Boulder in 1998, the M.A. degree in Music, Science, and Technology, at Stanford University, in 1999, the M.S. degree in multimedia engineering in the Media Arts and Technology program at University of California, Santa Barbara (UCSB), in 2004, and finally the M.S. and Ph.D. degrees in Electrical and Computer Engineering at UCSB, in 2007 and 2009, respectively. During 2009, Dr. Sturm was a Chateaubriand Post-doctoral Fellow at the Institut Jean Le Rond d’Alembert, Equipe Lutheries, Acoustique, Musique (LAM), at Université Pierre et Marie Curie, UPMC Paris 6.