Multilingually Trained Bottleneck Features in Spoken Language Recognition

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F17%3APU126448" target="_blank" >RIV/00216305:26230/17:PU126448 - isvavai.cz</a>

Výsledek na webu

<a href="http://www.sciencedirect.com/science/article/pii/S0885230816302947" target="_blank" >http://www.sciencedirect.com/science/article/pii/S0885230816302947</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1016/j.csl.2017.06.008" target="_blank" >10.1016/j.csl.2017.06.008</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Multilingually Trained Bottleneck Features in Spoken Language Recognition

Popis výsledku v původním jazyce

Multilingual training of neural networks has proven to be simple yet effective way to deal with multilingual training corpora. It allows to use several resources to jointly train a language independent representation of features, which can be encoded into low-dimensional feature set by embedding narrow bottleneck layer to the network. In this paper, we analyze such features on the task of spoken language recognition (SLR), focusing on practical aspects of training bottleneck networks and analyzing their integration in SLR. By comparing properties of mono and multilingual features we show the suitability of multilingual training for SLR. The state-of-the-art performance of these features is demonstrated on the NIST LRE09 database.

Název v anglickém jazyce

Multilingually Trained Bottleneck Features in Spoken Language Recognition

Popis výsledku anglicky

Multilingual training of neural networks has proven to be simple yet effective way to deal with multilingual training corpora. It allows to use several resources to jointly train a language independent representation of features, which can be encoded into low-dimensional feature set by embedding narrow bottleneck layer to the network. In this paper, we analyze such features on the task of spoken language recognition (SLR), focusing on practical aspects of training bottleneck networks and analyzing their integration in SLR. By comparing properties of mono and multilingual features we show the suitability of multilingual training for SLR. The state-of-the-art performance of these features is demonstrated on the NIST LRE09 database.

Druh

J_imp - Článek v periodiku v databázi Web of Science

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/LQ1602" target="_blank" >LQ1602: IT4Innovations excellence in science</a><br>

Návaznosti

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

Rok uplatnění

2017

Kód důvěrnosti údajů

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

Název periodika

COMPUTER SPEECH AND LANGUAGE

ISSN

0885-2308

e-ISSN

1095-8363

Svazek periodika

2017

Číslo periodika v rámci svazku

46

Stát vydavatele periodika

GB - Spojené království Velké Británie a Severního Irska

Počet stran výsledku

16

Strana od-do

252-267

Kód UT WoS článku

000407609600015

EID výsledku v databázi Scopus

2-s2.0-85023638945