GANSpaceSynth: A Hybrid Generative Adversarial Network Architecture for Organising the Latent Space using a Dimensionality Reduction for Real-Time Audio Synthesis

Abstract

Generative models enable possibilities in audio domain to present timbre as vectors in a high-dimensional latent space with Gen- erative Adversarial Networks (GANs). It is a common method in GAN models in which the musician’s control over timbre is mostly limited to sampling random points from the space and interpolating between them. In this paper, we present a novel hybrid GAN architecture that allows musicians to explore the GAN latent space in a more controlled manner, identifying the audio features in the trained checkpoints and giving an opportunity to specify particular audio features to be present or absent in the generated audio samples. We extend the paper with the detailed description of our GANSpaceSynth and present the Hallu composition tool as an application of this hybrid method in computer music practices.