Music transcription modelling and composition using deep learning

TL;DR

This paper explores the use of LSTM neural networks for music transcription modeling and composition, demonstrating their ability to generate new transcriptions that reflect specific musical styles and support creative processes.

Contribution

It introduces a practical LSTM-based system trained on a large dataset of ABC notation transcriptions, enabling music generation and analysis for composition and style reflection.

Findings

LSTM models can generate transcriptions resembling training data

Generated transcriptions reflect specific musical conventions

The system supports idea generation in music composition

Abstract

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}.