Coupled Recurrent Models for Polyphonic Music Composition

TL;DR

This paper presents a new recurrent neural model for polyphonic music composition that captures the structure of concurrent voices using a probabilistic factorization, combining convolutional and recurrent ideas.

Contribution

It introduces a novel conditional probabilistic factorization tailored for polyphonic music, integrating convolutional and recurrent neural networks for improved modeling.

Findings

Effective modeling of polyphonic scores with coupled recurrent models

Trained on 2,300 scores from the KernScores dataset

Demonstrates improved structure capturing in music generation

Abstract

This paper introduces a novel recurrent model for music composition that is tailored to the structure of polyphonic music. We propose an efficient new conditional probabilistic factorization of musical scores, viewing a score as a collection of concurrent, coupled sequences: i.e. voices. To model the conditional distributions, we borrow ideas from both convolutional and recurrent neural models; we argue that these ideas are natural for capturing music's pitch invariances, temporal structure, and polyphony. We train models for single-voice and multi-voice composition on 2,300 scores from the KernScores dataset.