Music transcription modelling and composition using deep learning
This work addresses music composition and transcription for musicians and researchers, but it is incremental as it applies existing deep learning methods to a specific domain.
The authors tackled music transcription and composition by applying LSTM networks to 23,000 transcriptions in ABC notation, generating new transcriptions and evaluating them through statistical comparisons, reflection of Celtic folk conventions, and use in idea generation for composition.
We apply deep learning methods, specifically long short-term memory (LSTM) networks, to music transcription modelling and composition. We build and train LSTM networks using approximately 23,000 music transcriptions expressed with a high-level vocabulary (ABC notation), and use them to generate new transcriptions. Our practical aim is to create music transcription models useful in particular contexts of music composition. We present results from three perspectives: 1) at the population level, comparing descriptive statistics of the set of training transcriptions and generated transcriptions; 2) at the individual level, examining how a generated transcription reflects the conventions of a music practice in the training transcriptions (Celtic folk); 3) at the application level, using the system for idea generation in music composition. We make our datasets, software and sound examples open and available: \url{https://github.com/IraKorshunova/folk-rnn}.