Projet de recherche doctoral numero :4789


Date depot: 1 janvier 1900
Titre: Music Interaction Design Through Co-Exploration
Directeur de thèse: Frédéric BEVILACQUA (STMS)
Domaine scientifique: Sciences et technologies de l'information et de la communication
Thématique CNRS : Non defini

Resumé: The main goal of this thesis is to design a general computational framework for music interaction design with a particular attention to assist user exploration along the design process in an adaptive way. Music interaction design requires technical skills in addition to creative insight. Interactive machine learning approaches have enabled more straightforward implementations by supporting custom motion-sound interaction design from direct demonstrations; however, these systems still lack of interactive assistance when it comes to fine-tuning a given design, or also to fostering new creative ideas from scratch. In this thesis, we argue that users should have the possibility to give feedback on the system's behavior, thus providing an additional space for expression and assistance in their own creative process. We believe machine learning algorithms have the computational abilities to support such real-time adaptation to users, as well as to direct design space search toward user-relevant information. We propose to formalize our approach as 'Music Interaction Design Through Co-Exploration', where co-exploration would stand for collaborative human exploration of design possibilities and computer exploration of parameter space. The project will be carried on in a interdisciplinary research context, implying the different fields in play in the UMR 9912 IRCAM-CNRS-UPMC, called Sciences and Technologies for Music and Sound. The scientific part of the project is centered around Information and Communication Technologies, including Computer Science (in particular Human-Computer Interaction and Artificial Intelligence/Machine Learning), Signal Processing, and Cognitive Science. The application context of the research is centered around music and sound technologies, from artistic applications to industrial applications in gaming or rehabilitation.

Doctorant.e: Scurto Hugo