Fast Multichannel NMF with Block-Diagonal Spatial Covariance Matrices for Efficient Blind Source Separation Using Distributed Microphone Arrays

TL;DR

The paper introduces a distributed FastMNMF approach with block-diagonal spatial covariance matrices, enabling efficient blind source separation across multiple microphone subarrays with reduced computation and improved performance.

Contribution

It proposes a novel distributed FastMNMF method that reduces computational cost by imposing block-diagonal structure on spatial covariance matrices, enabling scalable multi-microphone source separation.

Findings

Less computation time than conventional FastMNMF using all subarrays

Higher source-to-distortion ratio than single subarray FastMNMF

Effective in underdetermined multi-source scenarios

Abstract

Distributed microphone arrays composed of multiple subarrays enable blind source separation over a wide spatial area. Directly applying fast multichannel nonnegative matrix factorization (FastMNMF) to all subarrays can exploit observations from all subarrays, but it requires repeated inversions of large matrices spanning all microphones, causing the computational cost to increase rapidly as the number of microphones grows. In contrast, applying FastMNMF to one subarray reduces the matrix size but cannot exploit observations from other subarrays. We propose distributed FastMNMF, which imposes a block-diagonal structure on the source spatial covariance matrices, so that matrix inversions are performed within subarrays. The NMF-based source spectrogram model is shared across subarrays, allowing the method to aggregate source activity information while discarding inter-subarray covariance. In synchronized, noiseless simulations with fixed room and array/source geometry, the method required less computation time than conventional FastMNMF using all subarrays, achieved a higher average source-to-distortion ratio than conventional FastMNMF using one subarray, and was applicable in the tested five-source condition, where each four-microphone subarray was locally underdetermined.