Particle Filtering on the Audio Localization Manifold

TL;DR

This paper introduces a novel particle filtering algorithm that tracks moving sound sources by modeling the low-dimensional manifold of time difference of arrival (TDOA) delays, significantly improving audio localization accuracy.

Contribution

It combines manifold modeling with particle filtering and a new weighting scheme to enhance sound source tracking in high-dimensional delay spaces.

Findings

Outperforms standard particle filters in TDOA tracking accuracy

Effectively models low-dimensional manifolds of delays

Demonstrates robustness in audio localization tasks

Abstract

We present a novel particle filtering algorithm for tracking a moving sound source using a microphone array. If there are N microphones in the array, we track all $N \choose 2$ delays with a single particle filter over time. Since it is known that tracking in high dimensions is rife with difficulties, we instead integrate into our particle filter a model of the low dimensional manifold that these delays lie on. Our manifold model is based off of work on modeling low dimensional manifolds via random projection trees [1]. In addition, we also introduce a new weighting scheme to our particle filtering algorithm based on recent advancements in online learning. We show that our novel TDOA tracking algorithm that integrates a manifold model can greatly outperform standard particle filters on this audio tracking task.