Profile
Project title:
Explainability of AI Music Generation
Abstract:
The proposed research use explainable AI approaches to design and develop music generation systems that facilitate interactive and meaningful human-AI co-creation. The desired frameworks provide music producers with more controllable generative models to formulate their artistic intents, while providing effective human-AI communication channels that help the computational agent to pitch its creative decisions. The research will iteratively analyse, design, develop, and evaluate prototypes through user-centred approaches. The expected outcome will create novel and heuristic music production experiences that uniquely emerge in the intersection of AI and music, as well as a set of software frameworks ready to be deployed for virtual audio plug-ins.
C4DM theme affiliation:
Sonic Interaction Design