Unsupervised Acoustic Scene Mapping Based on Acoustic Features and Dimensionality Reduction

TL;DR

This paper presents an unsupervised deep learning approach using local conformal autoencoders to map acoustic scenes based on features, demonstrating robustness to reverberation and noise without relying on TDOA estimation.

Contribution

The paper introduces a novel unsupervised method for acoustic scene mapping that leverages local conformal autoencoders, avoiding traditional TDOA estimation and improving robustness to reverberation.

Findings

LOCA learns a spatially isometric representation of microphone locations.

The method outperforms other dimensionality reduction techniques in simulations.

LOCA shows significant robustness to reverberation effects.

Abstract

Classical methods for acoustic scene mapping require the estimation of time difference of arrival (TDOA) between microphones. Unfortunately, TDOA estimation is very sensitive to reverberation and additive noise. We introduce an unsupervised data-driven approach that exploits the natural structure of the data. Our method builds upon local conformal autoencoders (LOCA) - an offline deep learning scheme for learning standardized data coordinates from measurements. Our experimental setup includes a microphone array that measures the transmitted sound source at multiple locations across the acoustic enclosure. We demonstrate that LOCA learns a representation that is isometric to the spatial locations of the microphones. The performance of our method is evaluated using a series of realistic simulations and compared with other dimensionality-reduction schemes. We further assess the influence of reverberation on the results of LOCA and show that it demonstrates considerable robustness.