Text-based LSTM networks for Automatic Music Composition

TL;DR

This paper explores text-based LSTM neural networks for automatic music composition, demonstrating their ability to learn musical structures from text representations and offering tools for automatic or semi-automatic music creation.

Contribution

Introduces novel text-based LSTM methods for music composition, showing their effectiveness in learning musical relationships from text data.

Findings

Word-RNNs perform well in learning chord progressions and drum tracks.

Character-based RNNs succeed only in learning chord progressions.

The system supports automatic and semi-automatic music composition.

Abstract

In this paper, we introduce new methods and discuss results of text-based LSTM (Long Short-Term Memory) networks for automatic music composition. The proposed network is designed to learn relationships within text documents that represent chord progressions and drum tracks in two case studies. In the experiments, word-RNNs (Recurrent Neural Networks) show good results for both cases, while character-based RNNs (char-RNNs) only succeed to learn chord progressions. The proposed system can be used for fully automatic composition or as semi-automatic systems that help humans to compose music by controlling a diversity parameter of the model.