A cluster analysis of benchmark acoustic features on Brazilian music
In this work, we extend a standard and successful acoustic feature extraction approach based on trigger selection to examples of Brazilian Bossa-Nova and Heitor Villa Lobos music pieces. Additionally, we propose and implement a computational framework to disclose whether all the acoustic features extracted are statistically relevant, that is, non-redundant. Our experimental results show that not all these well-known features might be necessary for trigger selection, given the multivariate statistical redundancy found, which associated all these acoustic features into 3 clusters with different factor loadings and, consequently, representatives.
Andjela Markovic, Jürg Kühnis, and Lutz Jäncke. Task context influences brain activation during music listening. Frontiers in human neuroscience, 11:342, 2017.
Pasi Saari, Iballa Burunat, Elvira Brattico, and Petri Toiviainen. Decoding musical training from dynamic processing of musical features in the brain. Scientific Report, 708(8):1–12, 2018.
Peter Vuust, Elvira Brattico, Miia Sppanen, Risto Naatanen, and Mari Tervaniemi. The sound of music: Differentiating musicians using a fast, musical multi-feature mismatch negativity paradigm. Neuropsychologia, 50(7):1432–1443, June 2012.
C.A. Mikutta, G. Maissen, A. Altorfer, W. Strik, and T. Koenig. Professional musicians listen differently to music. Neuroscience, 268:102 – 111, 2014.
Vinoo Alluri, Petri Toiviainen, Iiro P Jääskeläinen, Enrico Glerean, Mikko Sams, and Elvira Brattico. Large-scale brain networks emerge from dynamic processing of musical timbre, key and rhythm. Neuroimage, 59(4):3677–3689, 2012.
Oliver Lartillot. MIRtoolbox 1.6.1 Users Manual, 2014.
Richard A. Johnson and Dean W. Wichern. Applied Multivariate Statistical Analysis. Pearson, 6th edition, 2007.