Weakly Supervised Audio Source Separation via Spectrum Energy Preserved Wasserstein Learning

TL;DR

This paper presents a novel weakly supervised deep learning method for audio source separation that uses Wasserstein distance and spectrum energy preservation, achieving competitive results without extensive prior assumptions.

Contribution

Introduces a spectrum energy preserved Wasserstein learning framework for weakly supervised audio source separation, reducing the need for prior model constraints.

Findings

Performs competitively on benchmark datasets.

Requires minimal prior model assumptions.

End-to-end training capability.

Abstract

Separating audio mixtures into individual instrument tracks has been a long standing challenging task. We introduce a novel weakly supervised audio source separation approach based on deep adversarial learning. Specifically, our loss function adopts the Wasserstein distance which directly measures the distribution distance between the separated sources and the real sources for each individual source. Moreover, a global regularization term is added to fulfill the spectrum energy preservation property regardless separation. Unlike state-of-the-art weakly supervised models which often involve deliberately devised constraints or careful model selection, our approach need little prior model specification on the data, and can be straightforwardly learned in an end-to-end fashion. We show that the proposed method performs competitively on public benchmark against state-of-the-art weakly supervised methods.