Learnable Sparse Filterbank for Speaker Verification

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F22%3APU146107" target="_blank" >RIV/00216305:26230/22:PU146107 - isvavai.cz</a>

Výsledek na webu

<a href="https://www.isca-speech.org/archive/pdfs/interspeech_2022/peng22e_interspeech.pdf" target="_blank" >https://www.isca-speech.org/archive/pdfs/interspeech_2022/peng22e_interspeech.pdf</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.21437/Interspeech.2022-11309" target="_blank" >10.21437/Interspeech.2022-11309</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Learnable Sparse Filterbank for Speaker Verification

Popis výsledku v původním jazyce

Recently, feature extraction with learnable filters was extensively investigated with speaker verification systems, with filters learned both in time- and frequency-domains. Most of the learned schemes however end up with filters close to their initialization (e.g. Mel filterbank) or filters strongly limited by their constraints. In this paper, we propose a novel learnable sparse filterbank, named LearnSF, by exclusively optimizing the sparsity of the filterbank, that does not explicitly constrain the filters to follow pre-defined distribution. After standard pre-processing (STFT and square of the magnitude spectrum), the learnable sparse filterbank is employed, with its normalized outputs fed into a neural network predicting the speaker identity. We evaluated the performance of the proposed approach on both VoxCeleb and CNCeleb datasets. The experimental results demonstrate the effectiveness of the proposed LearnSF compared to both widely-used acoustic features and existing parameterized learnable front-ends.

Název v anglickém jazyce

Learnable Sparse Filterbank for Speaker Verification

Popis výsledku anglicky

Recently, feature extraction with learnable filters was extensively investigated with speaker verification systems, with filters learned both in time- and frequency-domains. Most of the learned schemes however end up with filters close to their initialization (e.g. Mel filterbank) or filters strongly limited by their constraints. In this paper, we propose a novel learnable sparse filterbank, named LearnSF, by exclusively optimizing the sparsity of the filterbank, that does not explicitly constrain the filters to follow pre-defined distribution. After standard pre-processing (STFT and square of the magnitude spectrum), the learnable sparse filterbank is employed, with its normalized outputs fed into a neural network predicting the speaker identity. We evaluated the performance of the proposed approach on both VoxCeleb and CNCeleb datasets. The experimental results demonstrate the effectiveness of the proposed LearnSF compared to both widely-used acoustic features and existing parameterized learnable front-ends.

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

Výsledek vznikl pri realizaci vícero projektů. Více informací v záložce Projekty.

Návaznosti

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

Rok uplatnění

2022

Kód důvěrnosti údajů

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

Název statě ve sborníku

Proceedings of the Annual Conference of the International Speech Communication Association, INTERSPEECH

ISBN

—

ISSN

1990-9772

e-ISSN

—

Počet stran výsledku

5

Strana od-do

5110-5114

Název nakladatele

International Speech Communication Association

Místo vydání

Incheon

Místo konání akce

Incheon Korea

Datum konání akce

18. 9. 2022

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—