The project deals with multimodal interaction research aiming to identify behavioural indicators of depression in psychotherapeutic settings. Depression is a serious condition that affects over 300 million people globally and is linked to disability and mortality. A key factor for improvement in psychotherapy is the establishment of reciprocal and mutual interaction between client and therapist. Music therapy interventions can be used to stimulate spoken, musical and bodily interaction.
The ultimate goal of the project is to automatically detect depression and therapeutic success directly from temporal interaction patterns. The material used will be audiovisual recordings of clients with depression and their therapists during music therapy sessions. Video recordings of clients' and therapists' gestures, and audio recordings of conversations and musical improvisations will be analyzed. The research will take place at the Faculty of Information Technology of the University of Jyväskylä.