Experimenting with Musically Motivated Convolutional Neural Networks

TitleExperimenting with Musically Motivated Convolutional Neural Networks
Publication TypeConference Paper
Year of Publication2016
Conference Name14th International Workshop on Content-based Multimedia Indexing (CBMI 2016)
AuthorsPons, J., Lidy T., & Serra X.
Conference Start Date15/06/2016
PublisherIEEE
Conference LocationBucharest, Romania
KeywordsBallroom, convolutional neural networks, deep learning, Music
AbstractA common criticism of deep learning relates to the difficulty in understanding the underlying relationships that the neural networks are learning, thus behaving like a blackbox. In this article we explore various architectural choices of relevance for music signals classification tasks in order to start understanding what the chosen networks are learning. We first discuss how convolutional filters with different shapes can fit specific musical concepts and based on that we propose several musically motivated architectures. These architectures are then assessed by measuring the accuracy of the deep learning model in the prediction of various music classes using a known dataset of audio recordings of ballroom music. The classes in this dataset have a strong correlation with tempo, what allows assessing if the proposed architectures are learning frequency and/or time dependencies. Additionally, a black-box model is proposed as a baseline for comparison. With these experiments we have been able to understand what some deep learning based algorithms can learn from a particular set of data.
preprint/postprint documenthttps://repositori.upf.edu/handle/10230/27038
Additional material: 

 

intranet