Music Automaton (Musik-Automat)

The goal of the "Musik-Automat" project is to develop an AI-based app that supports musicians in composing. Musicians will be encouraged to create ideas through an interactive dialogue specifically designed for musical material.

The focus is on researching new methods of music data processing using neural networks and creating a powerful, interesting, and educational app. It will be implemented in a web browser, allowing users to interactively enter and "live" compose music, enabling a dialogue with the machine – almost like a game where humans and machines can start, continue, and end each other's musical phrases.

Andrew McLeod will bring its expertise in deep learning techniques for music analysis and generation to the project. This includes advice on model design and training methods, as well as input and output data representation. We will also bring access to appropriate sets of music data, as well as expertise and advice on building the front-end for the interactive game.

• Prof. Jörn Arnecke, Weimar University of Music FRANZ LISZT Weimar (project lead)
• Alex Vaughan, Weimar University of Music FRANZ LISZT Weimar

• Weimar University of Music FRANZ LISZT Weimar

October 2023 – December 2024