Εργαστήριο Μουσικής, Γνωσιακών Επιστημών και Κοινότητας
Προσκεκλημένη Ομιλία
Τετάρτη 18 Ιουνίου, 3 - 5μμ
Αίθουσα 917
Τμήμα Μουσικών Σπουδών, Φιλοσοφική Σχολή ΕΚΠΑ
Dr. Bastian Vobig
Technical University of Wuertzburg
Theoretical Background and Development of a Computational system for the Assessment of Musical Interaction in clinical Improvisation (CAMII)
In recent years, several interdisciplinary research efforts have aimed to integrate music
therapy, computer science, and music psychology to design and implement digital tools for
clinical practice in music therapy (Agres et al., 2021; Amorós-Sanchez et al., 2024; Wosch & Magee, 2018). Improvisational music therapy, in particular, benefits from digital processing
due to the complexity and richness of the musical and interpersonal data it generates. A
recurring challenge in employing these data within computational frameworks lies in
translating therapists’ high-level reasoning into low-level computational features—and vice
versa—to enable digital systems to provide clinically meaningful information (Volk et al.,
2023).
Over the past three years, a tool has been developed for the Computational Assessment of
Musical Interaction in Clinical Improvisation (CAMII), based on an interdisciplinary model
(Vobig, 2025). This model combines microanalytic methods of music therapy (Improvisation
Assessment Profile – Autonomy Microanalysis), digital tools for the analysis of clinical
improvisation (Music Therapy Toolbox), and sociological models (Social Systems Game
Theory). Using supervised machine learning, CAMII analyses MIDI data from dyadic piano
improvisations conducted in outpatient depression treatment and synthesises this data to
identify different types of musical interaction to characterise interpersonally related
individual musical behaviour of both client and therapist.
This presentation provides insights into the theoretical background and development of the
CAMII system, demonstrating how interdisciplinary methods can be synthesised to design a
computational framework for high-level analysis and assessment. The processes and
preliminary results of the prototypical system are illustrated using an example of musical
improvisation.