Challenging margin-based speaker embedding extractors by using the variational information bottleneck

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F24%3APU154968" target="_blank" >RIV/00216305:26230/24:PU154968 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.isca-archive.org/interspeech_2024/stafylakis24_interspeech.pdf" target="_blank" >https://www.isca-archive.org/interspeech_2024/stafylakis24_interspeech.pdf</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.21437/Interspeech.2024-2058" target="_blank" >10.21437/Interspeech.2024-2058</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Challenging margin-based speaker embedding extractors by using the variational information bottleneck

Popis výsledku v původním jazyce

Speaker embedding extractors are typically trained using a classification loss over the training speakers. During the last few years, the standard softmax/cross-entropy loss has been replaced by the margin-based losses, yielding significant im- provements in speaker recognition accuracy. Motivated by the fact that the margin merely reduces the logit of the target speaker during training, we consider a probabilistic framework that has a similar effect. The variational information bottle- neck provides a principled mechanism for making deterministic nodes stochastic, resulting in an implicit reduction of the pos- terior of the target speaker. We experiment with a wide range of speaker recognition benchmarks and scoring methods and re- port competitive results to those obtained with the state-of-the- art Additive Angular Margin loss.

Název v anglickém jazyce

Challenging margin-based speaker embedding extractors by using the variational information bottleneck

Popis výsledku anglicky

Speaker embedding extractors are typically trained using a classification loss over the training speakers. During the last few years, the standard softmax/cross-entropy loss has been replaced by the margin-based losses, yielding significant im- provements in speaker recognition accuracy. Motivated by the fact that the margin merely reduces the logit of the target speaker during training, we consider a probabilistic framework that has a similar effect. The variational information bottle- neck provides a principled mechanism for making deterministic nodes stochastic, resulting in an implicit reduction of the pos- terior of the target speaker. We experiment with a wide range of speaker recognition benchmarks and scoring methods and re- port competitive results to those obtained with the state-of-the- art Additive Angular Margin loss.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/VB02000060" target="_blank" >VB02000060: Nástroje boje proti hlasovým DeepFakes</a><br>

Návaznosti

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Rok uplatnění

2024

Kód důvěrnosti údajů

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Název statě ve sborníku

Proceedings of Interspeech 2024

ISBN

—

ISSN

1990-9772

e-ISSN

—

Počet stran výsledku

5

Strana od-do

3220-3224

Název nakladatele

International Speech Communication Association

Místo vydání

Kos

Místo konání akce

Kos

Datum konání akce

1. 9. 2024

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—