On the Use of Deep Mask Estimation Module for Neural Source Separation Systems

TL;DR

This paper investigates the impact of deep mask estimation modules in neural source separation systems, demonstrating that they effectively approximate the overseparation-grouping paradigm and improve performance over shallow modules.

Contribution

The paper provides an analysis connecting deep mask estimation modules to the overseparation-grouping paradigm, highlighting their efficiency in neural source separation.

Findings

Deep mask estimation modules outperform shallow ones.

Deep modules approximate overseparation-grouping paradigm.

Empirical validation shows improved source separation performance.

Abstract

Most of the recent neural source separation systems rely on a masking-based pipeline where a set of multiplicative masks are estimated from and applied to a signal representation of the input mixture. The estimation of such masks, in almost all network architectures, is done by a single layer followed by an optional nonlinear activation function. However, recent literatures have investigated the use of a deep mask estimation module and observed performance improvement compared to a shallow mask estimation module. In this paper, we analyze the role of such deeper mask estimation module by connecting it to a recently proposed unsupervised source separation method, and empirically show that the deep mask estimation module is an efficient approximation of the so-called overseparation-grouping paradigm with the conventional shallow mask estimation layers.