Saved in:
Bibliographic Details
Main Authors: Berg, Axel, Engman, Johanna, Gulin, Jens, Åström, Karl, Oskarsson, Magnus
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2408.17166
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916375580114944
author Berg, Axel
Engman, Johanna
Gulin, Jens
Åström, Karl
Oskarsson, Magnus
author_facet Berg, Axel
Engman, Johanna
Gulin, Jens
Åström, Karl
Oskarsson, Magnus
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2408_17166
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Learning Multi-Target TDOA Features for Sound Event Localization and Detection
Berg, Axel
Engman, Johanna
Gulin, Jens
Åström, Karl
Oskarsson, Magnus
Audio and Speech Processing
Machine Learning
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.
title Learning Multi-Target TDOA Features for Sound Event Localization and Detection
topic Audio and Speech Processing
Machine Learning
url https://arxiv.org/abs/2408.17166