Computational Musicology

This theme investigates musicological questions, in an empirical way, through the use both of algorithms and machine learning (in particular generative AI).. One strand of work seeks to understand music and the creative process through looking for stylistically successful ways of generating music from a corpus. Another strand looks at tools for supporting novices, whilst also again investigating creativity.

Key publications

• tiNNbre: A Timbre-based Musical Agent
• Sketching Music: exploring melodic similarity and contrast using a Digital Tabletop
• Generating Subjects for Pieces in the Style of Bach’s Two-Part Inventions

Resources

• A short video on Open Melody, a human-AI collaborative creative music tool
• Open Melody Web app