DNN Based Embeddings for Language Recognition

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216305%3A26230%2F18%3APU130737" target="_blank" >RIV/00216305:26230/18:PU130737 - isvavai.cz</a>

Výsledek na webu

<a href="http://www.fit.vutbr.cz/research/pubs/all.php?id=11723" target="_blank" >http://www.fit.vutbr.cz/research/pubs/all.php?id=11723</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/ICASSP.2018.8462403" target="_blank" >10.1109/ICASSP.2018.8462403</a>

Jazyk výsledku

angličtina

Název v původním jazyce

DNN Based Embeddings for Language Recognition

Popis výsledku v původním jazyce

In this work, we present a language identification (LID) system based on embeddings. In our case, an embedding is a fixed-length vector (similar to i-vector) that represents the whole utterance, but unlike i-vector it is designed to contain mostly information relevant to the target task (LID). In order to obtain these embeddings, we train a deep neural network (DNN) with sequence summarization layer to classify languages. In particular, we trained a DNN based on bidirectional long short-term memory (BLSTM) recurrent neural network (RNN) layers, whose frame-by-frame outputs are summarized into mean and standard deviation statistics. After this pooling layer, we add two fully connected layers whose outputs correspond to embeddings. Finally, we add a softmax output layer and train the whole network with multi-class cross-entropy objective to discriminate between languages. We report our results on NIST LRE 2015 and we compare the performance of embeddings and corresponding i-vectors both modeled by Gaussian Linear Classifier (GLC). Using only embeddings resulted in comparable performance to i-vectors and by performing score-level fusion we achieved 7.3% relative improvement over the baseline.

Název v anglickém jazyce

DNN Based Embeddings for Language Recognition

Popis výsledku anglicky

In this work, we present a language identification (LID) system based on embeddings. In our case, an embedding is a fixed-length vector (similar to i-vector) that represents the whole utterance, but unlike i-vector it is designed to contain mostly information relevant to the target task (LID). In order to obtain these embeddings, we train a deep neural network (DNN) with sequence summarization layer to classify languages. In particular, we trained a DNN based on bidirectional long short-term memory (BLSTM) recurrent neural network (RNN) layers, whose frame-by-frame outputs are summarized into mean and standard deviation statistics. After this pooling layer, we add two fully connected layers whose outputs correspond to embeddings. Finally, we add a softmax output layer and train the whole network with multi-class cross-entropy objective to discriminate between languages. We report our results on NIST LRE 2015 and we compare the performance of embeddings and corresponding i-vectors both modeled by Gaussian Linear Classifier (GLC). Using only embeddings resulted in comparable performance to i-vectors and by performing score-level fusion we achieved 7.3% relative improvement over the baseline.

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

2018

Kód důvěrnosti údajů

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

Název statě ve sborníku

Proceedings of ICASSP 2018

ISBN

978-1-5386-4658-8

ISSN

—

e-ISSN

—

Počet stran výsledku

5

Strana od-do

5184-5188

Název nakladatele

IEEE Signal Processing Society

Místo vydání

Calgary

Místo konání akce

Calgary

Datum konání akce

15. 4. 2018

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000446384605071