Conference paper Open Access
Parke-Wolfe, Samuel Thompson; Scurto, Hugo; Fiebrink, Rebecca
We have built a new software toolkit that enables music therapists and teachers to create custom digital musical interfaces for children with diverse disabilities. It was designed in collaboration with music therapists, teachers, and children. It uses interactive machine learning to create new sensor- and vision-based musical interfaces using demonstrations of actions and sound, making interface building fast and accessible to people without programming or engineering expertise. Interviews with two music therapy and education professionals who have used the software extensively illustrate how richly customised, sensor-based interfaces can be used in music therapy contexts; they also reveal how properties of input devices, music-making approaches, and mapping techniques can support a variety of interaction styles and therapy goals.