Multichannel Audio Source Separation with Independent Deeply Learned Matrix Analysis Using Product of Source Models

TL;DR

This paper introduces a novel multichannel audio source separation method that combines deep neural network-based and NMF-based source models using a product-of-expert approach, improving robustness to timbral mismatches.

Contribution

It extends IDLMA by integrating NMF-based source models with deep learning models through the product-of-source-models framework, enhancing separation performance.

Findings

Effective in separating sources with mismatched timbres

Improved separation accuracy over traditional IDLMA

Computationally efficient parameter estimation algorithm

Abstract

Independent deeply learned matrix analysis (IDLMA) is one of the state-of-the-art multichannel audio source separation methods using the source power estimation based on deep neural networks (DNNs). The DNN-based power estimation works well for sounds having timbres similar to the DNN training data. However, the sounds to which IDLMA is applied do not always have such timbres, and the timbral mismatch causes the performance degradation of IDLMA. To tackle this problem, we focus on a blind source separation counterpart of IDLMA, independent low-rank matrix analysis. It uses nonnegative matrix factorization (NMF) as the source model, which can capture source spectral components that only appear in the target mixture, using the low-rank structure of the source spectrogram as a clue. We thus extend the DNN-based source model to encompass the NMF-based source model on the basis of the product-of-expert concept, which we call the product of source models (PoSM). For the proposed PoSM-based IDLMA, we derive a computationally efficient parameter estimation algorithm based on an optimization principle called the majorization-minimization algorithm. Experimental evaluations show the effectiveness of the proposed method.