The Mirrornet : Learning Audio Synthesizer Controls Inspired by Sensorimotor Interaction

TL;DR

This paper introduces the MirrorNet, a neural architecture inspired by sensorimotor interactions, capable of learning audio synthesizer controls in an unsupervised way to produce and generalize melodies from auditory spectrograms.

Contribution

The paper presents the novel application of the MirrorNet architecture to learn synthesizer controls from auditory data without supervision, demonstrating its ability to generalize across different melodies and synthesizers.

Findings

MirrorNet accurately discovers synthesizer parameters for original melodies.

It generalizes to unseen melodies and different synthesizers.

It can approximate complex piano renditions with different control sets.

Abstract

Experiments to understand the sensorimotor neural interactions in the human cortical speech system support the existence of a bidirectional flow of interactions between the auditory and motor regions. Their key function is to enable the brain to `learn' how to control the vocal tract for speech production. This idea is the impetus for the recently proposed "MirrorNet", a constrained autoencoder architecture. In this paper, the MirrorNet is applied to learn, in an unsupervised manner, the controls of a specific audio synthesizer (DIVA) to produce melodies only from their auditory spectrograms. The results demonstrate how the MirrorNet discovers the synthesizer parameters to generate the melodies that closely resemble the original and those of unseen melodies, and even determine the best set parameters to approximate renditions of complex piano melodies generated by a different synthesizer. This generalizability of the MirrorNet illustrates its potential to discover from sensory data the controls of arbitrary motor-plants.