NIME2014-ESN.pdf (1.05 MB)
Musical instrument mapping design with Echo State Networks
Echo State Networks (ESNs), a form of recurrent neural network developed in the field of Reservoir Computing, show significant potential for use as a tool in the design of mappings for digital musical instruments. They have, however, seldom been used in this area, so this paper explores their possible applications. This project contributes a new open source library, which was developed to allow ESNs to run in the Pure Data dataflow environment. Several use cases were explored, focusing on addressing current issues in mapping research. ESNs were found to work successfully in scenarios of pattern classification, multiparametric control, explorative mapping and the design of nonlinearities and uncontrol. 'Un-trained' behaviours are proposed, as augmentations to the conventional reservoir system that allow the player to introduce potentially interesting non-linearities and uncontrol into the reservoir. Interactive evolution style controls are proposed as strategies to help design these behaviours, which are otherwise dependent on arbitrary values and coarse global controls. A study on sound classification showed that ESNs could reliably differentiate between two drum sounds, and also generalise to other similar input. Following evaluation of the use cases, heuristics are proposed to aid the use of ESNs in computer music scenarios.
History
Publication status
- Published
File Version
- Published version
Journal
Proceedings of the International Conference on New Interfaces for Musical ExpressionPublisher
NIMEPublisher URL
Page range
293-298Presentation Type
- paper
Event name
14th International Conference on New Interfaces for Musical ExpressionEvent location
Goldsmiths, University of LondonEvent type
conferenceEvent date
30 June - 4 July 2014Department affiliated with
- Informatics Publications
Full text available
- Yes
Peer reviewed?
- Yes
Legacy Posted Date
2015-01-05First Open Access (FOA) Date
2015-01-05First Compliant Deposit (FCD) Date
2014-12-28Usage metrics
Categories
No categories selectedKeywords
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC