Gesture-Based Musical Learning

Abstract

We give an overview of machine learning (ML) methods and how they can be used in interactive music and the design of new digital instruments. First, we introduce classification and regression, two ML tasks that are particularly relevant to gestural interaction, to the reader who is not a specialist. After that, we go over the existing NIME research that employs ML for musical gesture analysis and gestural sound control. We explain why live music performance is a relevant and challenging use case for machine learning and how machine learning can be used to create expressive musical interaction.