ABSTRACT
Musical agents are artificial agents that tackle musical creative tasks, partially or completely. This review of musical agents combines the terminology of Generative Arts (artistic practice) and the scientific literature of Computational Creativity, Multi-Agent Systems (MAS), and Artificial Intelligence. We define Musical Metacreation as a field that studies the partial or complete automation of musical tasks. We survey seventy-eight musical agent systems, and present a typology of musical agents. After examining the evaluation methodologies of musical agents, we propose possible future steps while mentioning ongoing discussions in the field.
Acknowledgements
We thank the anonymous reviewers for their insights and useful comments. We would like to thank our colleagues Ronald Boersen, Mirjana Prpa, Jianyu Fan, and Cale Plut for their suggestions and proofreading this paper. We also thank Ronald Boersen for helping us with Figure .
Disclosure statement
No potential conflict of interest was reported by the authors.
ORCID
Kıvanç Tatar http://orcid.org/0000-0003-4133-8641
Notes
1 Please refer to Churchill, and Prevost et al. (Citation2000) and Hartholt et al. (Citation2013) for an introduction to virtual agents in CGI.
2 Bretan Weinberg (Citation2016) present a state of the art in musically creative robotic agents.
4 Herremans, Chuan, and Chew (Citation2017) recently surveyed purely generative systems from a conventional music perspective.
5 Musical examples of OMAX in practice can be found at http://repmus.ircam.fr/omax/home
6 Wooldridge (Citation2009) provides the psuedo code of the BDI agent control loop.
7 One of the most famous breakbeat sections is the Amen Break, which is a 4 bar drum break of the song ‘Amen, Brother’ by 1960s soul/funk band the Winstons. The recording of Amen Break can be found at http://en.wikipedia.org/wiki/Amen_break.
8 MPEG 7 is a standardisation of low-level feature calculation and thumbnailling for multimedia.
9 Given a root note, Plomp-Levelt curves were proposed to calculate consonance or dissonance of any other note to the root note (Plomp Levelt, Citation1965).
11 The details of Sarva is available in the book on reinforcement learning by Sutton and Barto (Citation1998).