Learning Interpretable Representation for Controllable Polyphonic Music Generation
This work addresses the challenge of interpretable control in algorithmic composition for music generation, representing an incremental improvement through a novel method for a known bottleneck.
The paper tackles the problem of controlling polyphonic music generation by learning interpretable latent factors for chord and texture within a VAE framework, achieving successful disentanglement and high-quality controlled generation as shown by objective and subjective evaluations.
While deep generative models have become the leading methods for algorithmic composition, it remains a challenging problem to control the generation process because the latent variables of most deep-learning models lack good interpretability. Inspired by the content-style disentanglement idea, we design a novel architecture, under the VAE framework, that effectively learns two interpretable latent factors of polyphonic music: chord and texture. The current model focuses on learning 8-beat long piano composition segments. We show that such chord-texture disentanglement provides a controllable generation pathway leading to a wide spectrum of applications, including compositional style transfer, texture variation, and accompaniment arrangement. Both objective and subjective evaluations show that our method achieves a successful disentanglement and high quality controlled music generation.