# A Unified Neural Architecture for Instrumental Audio Tasks

## TL;DR

This paper introduces a unified neural architecture combining cGANs and WaveNet to perform multiple MIR tasks like pitch-tracking, source separation, and synthesis, demonstrating versatility and pioneering GAN application in instrument synthesis.

## Contribution

It presents the first end-to-end architecture that unifies various MIR tasks using cGANs and WaveNet, enabling efficient transfer learning and generalization across tasks.

## Key findings

- Successful application of cGANs for spectrogram translation across tasks
- Effective end-to-end model for multiple MIR tasks
- First use of GANs for guided instrument synthesis

## Abstract

Within Music Information Retrieval (MIR), prominent tasks -- including pitch-tracking, source-separation, super-resolution, and synthesis -- typically call for specialised methods, despite their similarities. Conditional Generative Adversarial Networks (cGANs) have been shown to be highly versatile in learning general image-to-image translations, but have not yet been adapted across MIR. In this work, we present an end-to-end supervisable architecture to perform all aforementioned audio tasks, consisting of a WaveNet synthesiser conditioned on the output of a jointly-trained cGAN spectrogram translator. In doing so, we demonstrate the potential of such flexible techniques to unify MIR tasks, promote efficient transfer learning, and converge research to the improvement of powerful, general methods. Finally, to the best of our knowledge, we present the first application of GANs to guided instrument synthesis.