Real-Time Musical Collaboration With a Probabilistic Model
Karl Johannsson, Victor Shepardson, Enrique Hurtado, Thor Magnusson, Hannes Högni Vilhjálmsson · 2024
This paper describes the development of a system to play an ancient Basque collaborative percussion instrument (the txalaparta) with a human performer. A deep probabilistic model was trained on two large datasets, and transfer learning used to train the model on high-quality data abstracted from txalaparta performances. The aim is to create a system that effectively responds to the human player in a performance environment, where the model's response must feel instantaneous, organic, and sensible in response to the user's input. The paper outlines the design and implementation of the system, including data collection, model training, and performance visualization. The study contributes to the field of interactive music systems and demonstrates the potential of deep learning in creating intelligent musical systems that can collaborate with human performers.