This project evaluates in the form of a pilot study the transfer and usage of a selected set of machine learning algorithms within compositional practice.
Recent developments in Machine Learning and Artificial Intelligence have the potential to foster new creative approaches within the field of computer assisted composition. This potential has so far been largely unexplored by composers due to a combination of required technical expertise and a lack of application scenarios and tools that are geared towards creative workflows.
This project evaluates in the form of a pilot study the transfer and usage of a selected set of machine learning algorithms within compositional practice. In particular, we address the supportive role that machine learning can play during the phases of ideation, exploration, and discovery of a creative workflow. In this pilot study, we focus on corpus-based compositional approaches that rely on large libraries of audio samples. For these approaches, we adopt machine learning algorithms and integrate them into software tools that aid composers in the navigation of existing and the creation of new audio material. The adoption of these tools in compositional practice is evaluated by inviting both composition students and accomplished composers to realize new musical works.
The outcome of the project includes gained knowledge concerning the creative potential of Machine Learning and Artificial Intelligence techniques for musical composition, software tools for the integration of these techniques within compositional practice, and new compositions that exemplify the creative potential of these techniques.
This project is conducted as a collaboration between guest researcher Kıvanç Tatar from the School of Interactive Arts and Technology, Simon Fraser University (Vancouver, BC, Canada) and members of the Institute of Computer Music and Sound Technologies, Zurich University of Arts. The invited researcher contributes to this project his expertise in musical applications of Machine Learning and Artificial Intelligence. The members of the host institution contribute their proficiency in transferring algorithmic and generative methods into tools for musical composition.