Computationally-engaged approaches to rhythm and musical heritage: Establishing cross-cultural relationships using data-driven approaches (2019-2021)

Funding: NYUAD Research Enhancement Fund ($249,634.00).

Principal Investigator: Carlos Guedes (NYUAD).

Co-Investigators: Andrew Eisenberg (NYUAD), Yi Fang (NYUAD), Beth Russell (NYUAD), Brian McFee (NYUNY) and Robert Rowe (NYUNY).

Post-Doctoral Associate: Kaustuv Ganguli (NYUAD).

Collaborators: Rebecca Pittam (NYUAD), Brad Bauer (NYUAD), Akshay Anantapadmanabhan (Independent musician), and Virginia Danielson (Independent researcher).

In this project, we intend to deepen the research undertaken in the previous project for the cross-cultural comparison of the music from the region, focusing this time more deeply in the music from the UAE, and on the development of generative music applications that can reliably produce stylistically-appropriate music derived from our data-driven analysis. Specifically, we will continue developing multidimensional data analysis and visualization approaches to the digital music collections we create and/or have access to, and consolidate our methodology for the creation of generative music applications. This will be done in four interconnected streams: 1) Improvement of algorithms and machine learning techniques for the computational analysis and understanding of music from the region; 2) Development of machine learning techniques for musical data interpretation; 3) Creation of a digital corpus of Emirati music consisting of existing recordings (to be digitized if necessary), and also of original recording of existing groups currently preserving the musical traditions of the Emirates; 4) Creation of real-time generative music models.

We believe the work proposed here has the potential to advance the state of knowledge related to the collection and preservation of musical styles from the region while opening that knowledge to new analytical techniques, namely through the development of machine-learning analysis and visualization strategies of the multidimensional aspects of this data. This will certainly lead to new insights about relationships between the different styles from this rather vast region. It will underpin advances in modeling our understanding of the human experience of rhythm, and consequently of music, as well as permit novel creative processes based on such models for the generation and dissemination of interactive artworks and applications aimed at a broad public.