A relevance score estimation for spoken term detection based on RNN-generated pronunciation embeddings

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F49777513%3A23520%2F17%3A43932653" target="_blank" >RIV/49777513:23520/17:43932653 - isvavai.cz</a>

Výsledek na webu

<a href="https://pdfs.semanticscholar.org/a8ad/654be9b7b1c3914ac69a697850fc4657473b.pdf" target="_blank" >https://pdfs.semanticscholar.org/a8ad/654be9b7b1c3914ac69a697850fc4657473b.pdf</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.21437/Interspeech.2017-1087" target="_blank" >10.21437/Interspeech.2017-1087</a>

Jazyk výsledku

angličtina

Název v původním jazyce

A relevance score estimation for spoken term detection based on RNN-generated pronunciation embeddings

Popis výsledku v původním jazyce

In this paper, we present a novel method for term score es- timation. The method is primarily designed for scoring the out-of-vocabulary terms, however it could also estimate scores for in-vocabulary results. The term score is computed as a co- sine distance of two pronunciation embeddings. The first one is generated from the grapheme representation of the searched term, while the second one is computed from the recognized phoneme confusion network. The embeddings are generated by specifically trained recurrent neural network built on the idea of Siamese neural networks. The RNN is trained from recognition results on word- and phone-level in an unsupervised fashion without need of any hand-labeled data. The method is evaluated on the MALACH data in two languages, English and Czech. The results are compared with two baseline methods for OOV term detection.

Název v anglickém jazyce

A relevance score estimation for spoken term detection based on RNN-generated pronunciation embeddings

Popis výsledku anglicky

In this paper, we present a novel method for term score es- timation. The method is primarily designed for scoring the out-of-vocabulary terms, however it could also estimate scores for in-vocabulary results. The term score is computed as a co- sine distance of two pronunciation embeddings. The first one is generated from the grapheme representation of the searched term, while the second one is computed from the recognized phoneme confusion network. The embeddings are generated by specifically trained recurrent neural network built on the idea of Siamese neural networks. The RNN is trained from recognition results on word- and phone-level in an unsupervised fashion without need of any hand-labeled data. The method is evaluated on the MALACH data in two languages, English and Czech. The results are compared with two baseline methods for OOV term detection.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

20205 - Automation and control systems

Projekt

<a href="/cs/project/GBP103%2F12%2FG084" target="_blank" >GBP103/12/G084: Centrum pro multi-modální interpretaci dat velkého rozsahu</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 statě ve sborníku

Proceedings of the 18th Annual Conference of the International Speech Communication Association (Interspeech 2017)

ISBN

978-1-5108-4876-4

ISSN

1990-9772

e-ISSN

—

Počet stran výsledku

5

Strana od-do

2934-2938

Název nakladatele

Curran Associates, Inc.

Místo vydání

Red Hook, NY

Místo konání akce

Stockholm, Sweden

Datum konání akce

20. 8. 2017

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000457505000607