Abstract

Advances in the latest years on neural generative models such as GANs and VAEs have unveiled a great potential for creative applications supported by artificial intelligence methods. The most known applications have occurred in areas such as image synthesis for face generation as well as in natural language generation. In terms of tools for music composition, several systems have been released in the latest years, but there is still space for improving the possibilities of music co-creation with neural generative tools. In this context, we introduce Latent Chords, a system based on a Variational Autoencoder architecture which learns a latent space by reconstructing piano chords. We provide details of the neural architecture, the training process and we also show how Latent Chords can be used for a controllable exploration of chord sounds as well as to generate new chords by manipulating the latent representation.