@inproceedings{shen-etal-2014-nct,
title = "The {NCT} {ASR} system for {IWSLT} 2014",
author = "Shen, Peng and
Lu, Yugang and
Hu, Xinhui and
Kanda, Naoyuki and
Saiko, Masahiro and
Hori, Chiori",
booktitle = "Proceedings of the 11th International Workshop on Spoken Language Translation: Evaluation Campaign",
month = dec # " 4-5",
year = "2014",
address = "Lake Tahoe, California",
url = "https://aclanthology.org/2014.iwslt-evaluation.16",
pages = "113--118",
abstract = "This paper describes our automatic speech recognition system for IWSLT2014 evaluation campaign. The system is based on weighted finite-state transducers and a combination of multiple subsystems which consists of four types of acoustic feature sets, four types of acoustic models, and N-gram and recurrent neural network language models. Compared with our system used in last year, we added additional subsystems based on deep neural network modeling on filter bank feature and convolutional deep neural network modeling on filter bank feature with tonal features. In addition, modifications and improvements on automatic acoustic segmentation and deep neural network speaker adaptation were applied. Compared with our last year{'}s system on speech recognition experiments, our new system achieved 21.5{\%} relative improvement on word error rate on the 2013 English test data set.",
}
<?xml version="1.0" encoding="UTF-8"?>
<modsCollection xmlns="http://www.loc.gov/mods/v3">
<mods ID="shen-etal-2014-nct">
<titleInfo>
<title>The NCT ASR system for IWSLT 2014</title>
</titleInfo>
<name type="personal">
<namePart type="given">Peng</namePart>
<namePart type="family">Shen</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Yugang</namePart>
<namePart type="family">Lu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Xinhui</namePart>
<namePart type="family">Hu</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Naoyuki</namePart>
<namePart type="family">Kanda</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Masahiro</namePart>
<namePart type="family">Saiko</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<name type="personal">
<namePart type="given">Chiori</namePart>
<namePart type="family">Hori</namePart>
<role>
<roleTerm authority="marcrelator" type="text">author</roleTerm>
</role>
</name>
<originInfo>
<dateIssued>2014-dec" 4-5"</dateIssued>
</originInfo>
<typeOfResource>text</typeOfResource>
<relatedItem type="host">
<titleInfo>
<title>Proceedings of the 11th International Workshop on Spoken Language Translation: Evaluation Campaign</title>
</titleInfo>
<originInfo>
<place>
<placeTerm type="text">Lake Tahoe, California</placeTerm>
</place>
</originInfo>
<genre authority="marcgt">conference publication</genre>
</relatedItem>
<abstract>This paper describes our automatic speech recognition system for IWSLT2014 evaluation campaign. The system is based on weighted finite-state transducers and a combination of multiple subsystems which consists of four types of acoustic feature sets, four types of acoustic models, and N-gram and recurrent neural network language models. Compared with our system used in last year, we added additional subsystems based on deep neural network modeling on filter bank feature and convolutional deep neural network modeling on filter bank feature with tonal features. In addition, modifications and improvements on automatic acoustic segmentation and deep neural network speaker adaptation were applied. Compared with our last year’s system on speech recognition experiments, our new system achieved 21.5% relative improvement on word error rate on the 2013 English test data set.</abstract>
<identifier type="citekey">shen-etal-2014-nct</identifier>
<location>
<url>https://aclanthology.org/2014.iwslt-evaluation.16</url>
</location>
<part>
<date>2014-dec" 4-5"</date>
<extent unit="page">
<start>113</start>
<end>118</end>
</extent>
</part>
</mods>
</modsCollection>
%0 Conference Proceedings
%T The NCT ASR system for IWSLT 2014
%A Shen, Peng
%A Lu, Yugang
%A Hu, Xinhui
%A Kanda, Naoyuki
%A Saiko, Masahiro
%A Hori, Chiori
%S Proceedings of the 11th International Workshop on Spoken Language Translation: Evaluation Campaign
%D 2014
%8 dec" 4 5"
%C Lake Tahoe, California
%F shen-etal-2014-nct
%X This paper describes our automatic speech recognition system for IWSLT2014 evaluation campaign. The system is based on weighted finite-state transducers and a combination of multiple subsystems which consists of four types of acoustic feature sets, four types of acoustic models, and N-gram and recurrent neural network language models. Compared with our system used in last year, we added additional subsystems based on deep neural network modeling on filter bank feature and convolutional deep neural network modeling on filter bank feature with tonal features. In addition, modifications and improvements on automatic acoustic segmentation and deep neural network speaker adaptation were applied. Compared with our last year’s system on speech recognition experiments, our new system achieved 21.5% relative improvement on word error rate on the 2013 English test data set.
%U https://aclanthology.org/2014.iwslt-evaluation.16
%P 113-118
Markdown (Informal)
[The NCT ASR system for IWSLT 2014](https://aclanthology.org/2014.iwslt-evaluation.16) (Shen et al., IWSLT 2014)
ACL
- Peng Shen, Yugang Lu, Xinhui Hu, Naoyuki Kanda, Masahiro Saiko, and Chiori Hori. 2014. The NCT ASR system for IWSLT 2014. In Proceedings of the 11th International Workshop on Spoken Language Translation: Evaluation Campaign, pages 113–118, Lake Tahoe, California.