eduzhai > Applied Sciences > Engineering >

Modeling Baroque Two-Part Counterpoint with Neural Machine Translation

  • Save

... pages left unread,continue reading

Document pages: 5 pages

Abstract: We propose a system for contrapuntal music generation based on a NeuralMachine Translation (NMT) paradigm. We consider Baroque counterpoint and areinterested in modeling the interaction between any two given parts as a mappingbetween a given source material and an appropriate target material. Like intranslation, the former imposes some constraints on the latter, but doesn tdefine it completely. We collate and edit a bespoke dataset of Baroque pieces,use it to train an attention-based neural network model, and evaluate thegenerated output via BLEU score and musicological analysis. We show that ourmodel is able to respond with some idiomatic trademarks, such as imitation andappropriate rhythmic offset, although it falls short of having learnedstylistically correct contrapuntal motion (e.g., avoidance of parallel fifths)or stricter imitative rules, such as canon.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...