A Model Compression Method with Matrix Product Operators for Speech Enhancement

TL;DR

This paper introduces a matrix product operator (MPO) based model compression technique for neural networks in speech enhancement, significantly reducing parameters while outperforming pruning methods, especially at low compression rates.

Contribution

The paper presents a novel MPO-based compression method for DNNs in speech enhancement, replacing weight matrices with MPO format before training.

Findings

MPO-based method outperforms pruning in speech enhancement tasks.

Effective compression achieved at various rates, especially low compression.

Improved performance in resource-limited device applications.

Abstract

The deep neural network (DNN) based speech enhancement approaches have achieved promising performance. However, the number of parameters involved in these methods is usually enormous for the real applications of speech enhancement on the device with the limited resources. This seriously restricts the applications. To deal with this issue, model compression techniques are being widely studied. In this paper, we propose a model compression method based on matrix product operators (MPO) to substantially reduce the number of parameters in DNN models for speech enhancement. In this method, the weight matrices in the linear transformations of neural network model are replaced by the MPO decomposition format before training. In experiment, this process is applied to the causal neural network models, such as the feedforward multilayer perceptron (MLP) and long short-term memory (LSTM) models. Both MLP and LSTM models with/without compression are then utilized to estimate the ideal ratio mask for monaural speech enhancement. The experimental results show that our proposed MPO-based method outperforms the widely-used pruning method for speech enhancement under various compression rates, and further improvement can be achieved with respect to low compression rates. Our proposal provides an effective model compression method for speech enhancement, especially in cloud-free application.