Estimation of Microphone Clusters in Acoustic Sensor Networks using Unsupervised Federated Learning

TL;DR

This paper introduces a privacy-preserving unsupervised federated learning method using autoencoders to estimate microphone clusters in acoustic sensor networks, optimizing for scarce data and enabling effective cluster membership determination.

Contribution

It adapts clustered federated learning to unsupervised scenarios with a lightweight autoencoder, and introduces a method for computing cluster membership in acoustic sensor networks.

Findings

Effective clustering measured by clustering-based metrics.

Successful network-wide classification performance.

Robustness with very scarce training data.

Abstract

In this paper we present a privacy-aware method for estimating source-dominated microphone clusters in the context of acoustic sensor networks (ASNs). The approach is based on clustered federated learning which we adapt to unsupervised scenarios by employing a light-weight autoencoder model. The model is further optimized for training on very scarce data. In order to best harness the benefits of clustered microphone nodes in ASN applications, a method for the computation of cluster membership values is introduced. We validate the performance of the proposed approach using clustering-based measures and a network-wide classification task.