Audio representations for deep learning in sound synthesis: A review

TL;DR

This review paper discusses various audio representations used in deep learning-based sound synthesis, highlighting how different representations influence model architecture choices, training efficiency, and sound quality evaluation.

Contribution

It provides a comprehensive overview of audio representations and their impact on deep learning sound synthesis architectures and evaluation methods.

Findings

Different audio representations affect model complexity and training time.

Transformations like feature extraction improve efficiency and perceptual relevance.

Evaluation metrics vary depending on the audio representation used.

Abstract

The rise of deep learning algorithms has led many researchers to withdraw from using classic signal processing methods for sound generation. Deep learning models have achieved expressive voice synthesis, realistic sound textures, and musical notes from virtual instruments. However, the most suitable deep learning architecture is still under investigation. The choice of architecture is tightly coupled to the audio representations. A sound's original waveform can be too dense and rich for deep learning models to deal with efficiently - and complexity increases training time and computational cost. Also, it does not represent sound in the manner in which it is perceived. Therefore, in many cases, the raw audio has been transformed into a compressed and more meaningful form using upsampling, feature-extraction, or even by adopting a higher level illustration of the waveform. Furthermore, conditional on the form chosen, additional conditioning representations, different model architectures, and numerous metrics for evaluating the reconstructed sound have been investigated. This paper provides an overview of audio representations applied to sound synthesis using deep learning. Additionally, it presents the most significant methods for developing and evaluating a sound synthesis architecture using deep learning models, always depending on the audio representation.