Food & Paper: Connecting ethnomusicology and machine learning: Multidisciplinary analysis of Maracatu de Baque Solto (Baraldi and Davies)

This week's Food & Paper will be given by Filippo Bonini Baraldi (Universidade Nova) and Matthew Davies (University of Coimbra) on Connecting ethnomusicology and machine learning.

Filippo Bonini Baraldi and Matthew Davies


Maracatu de Baque Solto is a carnival practice in rural Pernambuco in north eastern Brazil, which is characterised by alternating periods of complex dance and loud, fast-paced music with improvised poetry. Maracatu is understood to bestow a special protective function on local communities and is thus strongly tied to a sense of collective well-being. Unlike many other musics from Brazil which are well-known globally, Maracatu remains a highly localised practice and has been the subject of very little academic research. Within the context of the Portuguese nationally-funded project “HELP-MD - The Healing and Emotional Power of Music and Dance”, we execute a strongly multidisciplinary approach to understanding the performative and aesthetic strategies within Maracatu, which links ethnomusicological field research with motion capture, audio engineering, music signal processing, and machine learning. In this presentation we seek to highlight the necessity of such a multidisciplinary approach and in turn how this can enable the observation of precise timing information which can lead to new ethnomusicological understanding of the practice of Maracatu.


Filippo Bonini Baraldi is an assistant researcher (Investigador FCT fellow, 2015-20) at the Instituto de Etnomusicologia (INET-MD), Universidade Nova, Lisbon (Portugal), and an associate member of the Centre de Recherche en Ethnomusicologie (CREM-LESC, Université Paris Ouest Nanterre, France). He obtained a MA degree ("Laurea") in Electronic Engineering at Padova University (Italy) in 2001, completed a MA degree in Music Technology at the Institute of Research and Coordination Acoustic/Music (IRCAM, Paris) in 2003, and accomplished a Ph.D. in Ethnomusicology at Nanterre University in 2010 with the jury's unanimously honors. He obtained a doctoral grant by the French National Center for Scientific Research (CNRS, 2003-2006) and a post-doctoral fellowship by the Society of Ethnology of Nanterre (E. Fleichman, 2011). His doctoral researches have been awarded the prestigious international Ph.D. thesis prize by the Quai Branly Museum (Paris, 2011). 

Matthew Davies is a music information retrieval researcher with a background in digital signal processing, and currently holds an FCT Investigator Development Grant at CISUC, Centre for Informatics and Systems of the University of Coimbra, and is an external collaborator at INESC TEC. His main research interests include the analysis of rhythm in musical audio signals, evaluation methodology, creative music applications, music therapy, and reproducible research. His most recent research has been in the development of compact and computationally efficient deep neural networks for the analysis of rhythmic structure. From 2014-2019, Matthew coordinated the Sound and Music Computing Group in the Centre for Telecommunications and Multimedia at INESC TEC, and was an Invited Assistant Professor at the Faculty of Engineering of the University of Porto, in Portugal. 

Published Jan. 27, 2021 1:45 PM - Last modified Jan. 27, 2021 1:45 PM