BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Tracking the odd: Meter inference in musical audio using particle 
 filters - Dr Andre Holzapfel\, Boğaziçi University\, Istanbul
DTSTART:20140630T130000Z
DTEND:20140630T140000Z
UID:TALK52869@talks.cam.ac.uk
CONTACT:Prof. Ramji Venkataramanan
DESCRIPTION:In the field of MIR (Music Information Retrieval) the potentia
 l of Bayesian methods for signal analysis is widely acknowledged\, but for
  many tasks in MIR Bayesian approaches are still the exception. In this ta
 lk a new state-of-the-art particle filter (PF) system is presented for the
  task of meter inference from a music audio signal. The system can be appl
 ied to determine the type of meter of a musical audio signal\, and to trac
 k beats and downbeats in a unified framework. The new inference method is 
 designed to overcome the problem of PFs in multi-modal state-spaces\, whic
 h arise due to tempo and phase ambiguities in musical rhythm representatio
 ns. Our recent work will be summarized in which we compare the new method 
 with a hidden Markov model (HMM) system and several other PF schemes in te
 rms of performance\, speed and scalability on several audio datasets. We d
 emonstrate that using the proposed system the computational complexity can
  be reduced drastically in comparison to the HMM while maintaining the sam
 e order of downbeat and beat tracking accuracy. Furthermore\, our results 
 on a culturally diverse collection will be illustrated\, in which the prop
 osed model is shown to infer the type of meter in a collection of music sa
 mples from India\, Greece\, and Turkey. Our results suggest that the propo
 sed system is capable of meter inference in large culturally diverse music
  collections. We argue that the system can be easily adapted to musical st
 yles\, and therefore avoids inclusion of an ethnocentric bias into music r
 ecommendation and distribution software systems. \n\n*BIO*: Andre Holzapfe
 l is currently a post-doctoral researcher at\nBoğaziçi University\, Ista
 nbul\, funded by a Marie-Curie IEF\ngrant. In parallel he pursues his seco
 nd doctoral degree in music at\nthe Center for Advanced Music Studies (MIA
 M)\, Istanbul. Before his\nwork in Istanbul he was working as a researcher
  for the CompMusic\nproject (University Pompeu Fabra)\, for INESC TEC Port
 o\, and for the\nAustrian Research Institute for Artificial Intelligence. 
 He obtained\nhis first Ph.D. in Computer Science at the University of Cret
 e. His\nMIR related research focuses on models and inference schemes for t
 he structure of music\, with an emphasis on rhythm. His research in\nethno
 musicology focuses on music of Crete\, and he investigates \nsubjects of t
 he interaction of music and technology. He is a regular\nperformer of Gree
 k Rembetiko music\, with his main instruments the\nTurkish oud and the gui
 tar. He is currently directing a documentary movie on amateur Fado in the 
 city of Porto.  For further information\, refer to "www.rhythmos.org":http
 ://www.rhythmos.org
LOCATION:BE4-38 (CBL Meeting Room)
END:VEVENT
END:VCALENDAR