Multi-channel Speech Separation Using Deep Embedding Model with Multilayer Bootstrap Networks

TL;DR

This paper introduces DPCL++, an improved deep clustering method for speech separation that employs multilayer bootstrap networks to enhance robustness in reverberant environments, especially when training and testing conditions differ.

Contribution

The paper proposes integrating multilayer bootstrap networks into deep clustering to reduce noise and variations in embeddings, improving speech separation in challenging environments.

Findings

Enhanced separation accuracy in reverberant environments

Robustness to environment mismatch demonstrated

Effective noise reduction in embedding vectors

Abstract

Recently, deep clustering (DPCL) based speaker-independent speech separation has drawn much attention, since it needs little speaker prior information. However, it still has much room of improvement, particularly in reverberant environments. If the training and test environments mismatch which is a common case, the embedding vectors produced by DPCL may contain much noise and many small variations. To deal with the problem, we propose a variant of DPCL, named DPCL++, by applying a recent unsupervised deep learning method---multilayer bootstrap networks(MBN)---to further reduce the noise and small variations of the embedding vectors in an unsupervised way in the test stage, which fascinates k-means to produce a good result. MBN builds a gradually narrowed network from bottom-up via a stack of k-centroids clustering ensembles, where the k-centroids clusterings are trained independently by random sampling and one-nearest-neighbor optimization. To further improve the robustness of DPCL++ in reverberant environments, we take spatial features as part of its input. Experimental results demonstrate the effectiveness of the proposed method.