Position Tracking of a Varying Number of Sound Sources with Sliding Permutation Invariant Training

Tutkimustuotos: KonferenssiartikkeliScientificvertaisarvioitu

2 Sitaatiot (Scopus)
1 Lataukset (Pure)

Abstrakti

Machine-learning-based sound source localization (SSL) methods have shown strong performance in challenging acoustic scenarios. However, little work has been done on adapting such methods to track consistently multiple sources appearing and disappearing, as would occur in reality. In this paper, we present a new training strategy for deep learning SSL models with a straightforward implementation based on the mean squared error of the optimal association between estimated and reference positions in the preceding time frames. It optimizes the desired properties of a tracking system: handling a time-varying number of sources and ordering localization estimates according to their trajectories, minimizing identity switches (IDSs). Evaluation on simulated data of multiple reverberant moving sources and on two model architectures proves its effectiveness in reducing identity switches without compromising frame-wise localization accuracy.
AlkuperäiskieliEnglanti
Otsikko2023 31st European Signal Processing Conference (EUSIPCO)
KustantajaIEEE
Sivut251-255
ISBN (elektroninen)978-9-4645-9360-0
DOI - pysyväislinkit
TilaJulkaistu - 4 syysk. 2023
OKM-julkaisutyyppiA4 Artikkeli konferenssijulkaisussa
TapahtumaEuropean Signal Processing Conference - Helsinki, Suomi
Kesto: 4 syysk. 20238 syysk. 2023

Julkaisusarja

NimiEuropean Signal Processing Conference
ISSN (elektroninen)2076-1465

Conference

ConferenceEuropean Signal Processing Conference
Maa/AlueSuomi
KaupunkiHelsinki
Ajanjakso4/09/238/09/23

Julkaisufoorumi-taso

  • Jufo-taso 1

Sormenjälki

Sukella tutkimusaiheisiin 'Position Tracking of a Varying Number of Sound Sources with Sliding Permutation Invariant Training'. Ne muodostavat yhdessä ainutlaatuisen sormenjäljen.

Siteeraa tätä