Adaptive Quantization for Multichannel Wiener Filter-Based Speech Enhancement in Wireless Acoustic Sensor Networks

Abstract

Speech enhancement in wireless acoustic sensor networks requires the exchange of audio signals. Since the wireless communication often dominates the nodes' energy budget, techniques for data exchange reduction are crucial. Adaptive quantization aims to optimize the bit depth of each exchanged signal according to its contribution to the speech enhancement performance. This enables the network to scale its energy and communication bandwidth requirements according to the current operating environment. The impact metric was previously proposed to predict the effect of quantization in linear minimum mean squared error (MMSE) estimation. We provide new insights into greedy adaptive quantization based on this impact metric. We achieve this by expanding the mathematical framework to include a new metric based on the gradient of the MMSE as a function of the quantization noise power. Using these tools, we show how the MMSE gradient naturally leads to a greedy algorithm and how the impact metric is a generalization of the gradient metric and a previously proposed metric. Besides, we validate the impact metric for adaptive quantization both in a simulated and in a real wireless acoustic sensor network deployed in a home environment, showing the energy savings achievable through greedy adaptive quantization.

Utility Metrics for Assessment and Subset Selection of Input Variables for Linear Estimation [Tips & Tricks]

“…The backward greedy algorithm can even be shown to be optimal if an exact or almost-exact sparse solution exists [1]. In a similar fashion, the noise impact metrics (19)-(20) can be used for, e.g., a greedy adaptive quantization, where in each iteration a certain amount of quantization noise is added to the input with the lowest noise-impact [11], [16]. Finally, a utility-based greedy variable selection based on (24) yields a combination of variable selection with 2-norm minimization, which is akin to the so-called elastic net procedure [17].…”

Model-Based Voice Activity Detection in Wireless Acoustic Sensor Networks

“…This section formulates the centralized detection problem. We apply the GLRT method to solve the VAD problem in (1). Based on the detection theory, the GLRT makes the decision with the following function:…”

“…Compared to the traditional microphone arrays, WASNs are more flexible and scalable, and are able to physically cover a larger space and capture more spatial information. The distributed speech enhancement methods, such as the distributed Wiener filter [1], the distributed maximum SNR filter [2], the distributed beamforming [3], need an estimate of the second-order statistics of the noise before forming the linear filter. Usually, the noise covariance matrix is estimated in a recursive way, and the estimated covariance matrix is updated only when the speech is absent.…”

Distributed Microphone Speech Enhancement based on Deep Learning

“…The received noises inevitably degrade the performance of multi-channel (MC)-based human-human and humanmachine interfaces, and this issue has attracted significant attention over the years [1,2,3]. In recent decades, numerous MC speechenhancement (SE) approaches have been proposed to alleviate the effect of noise and improve the quality and intelligibility [4,5,6,7] of received speech signals. In general, most of these approaches have been proposed for use in a microphone array architecture, wherein multiple microphones are compactly placed in a small space.…”