Learning Multi-Target TDOA Features for Sound Event Localization and Detection
Abstract
Sound event localization and detection (SELD) systems using audio recordings from a microphone array rely on spatial cues for determining the location of sound events. As a consequence, the localization performance of such systems is to a large extent determined by the quality of the audio features that are used as inputs to the system. We propose a new feature, based on neural generalized cross-correlations with phase-transform (NGCC-PHAT), that learns audio representations suitable for localization. Using permutation invariant training for the time-difference of arrival (TDOA) estimation problem enables NGCC-PHAT to learn TDOA features for multiple overlapping sound events. These features can be used as a drop-in replacement for GCC-PHAT inputs to a SELD-network. We test our method on the STARSS23 dataset and demonstrate improved localization performance compared to using standard GCC-PHAT or SALSA-Lite input features.
Keywords
Cite
@article{arxiv.2408.17166,
title = {Learning Multi-Target TDOA Features for Sound Event Localization and Detection},
author = {Axel Berg and Johanna Engman and Jens Gulin and Karl Åström and Magnus Oskarsson},
journal= {arXiv preprint arXiv:2408.17166},
year = {2024}
}
Comments
DCASE 2024